[PATCH V2] clk: bcm2835: mark enabled pll_dividers as critical

[Eric Anholt]

kernel at martin.sperl.org writes:

> From: Martin Sperl <kernel at martin.sperl.org>
>
> The bcm2835 firmware enables several clocks and plls before
> booting the linux kernel.
>
> These plls should never get disabled as it may result in a
> stopped system clock and more.
>
> So during probing we check if the pll_divider is enabled
> and if it is then mark that pll_divider as critical.
> As a consequence this will also enable the corresponding parent pll.
>
> Here the list of pll_div that are marked as critical:
> [    2.022437] bcm2835-clk 20101000.cprman: found enabled pll_div plla_core - marking it as critical
> [    2.031640] bcm2835-clk 20101000.cprman: found enabled pll_div pllb_arm - marking it as critical
> [    2.040966] bcm2835-clk 20101000.cprman: found enabled pll_div pllc_core0 - marking it as critical
> [    2.050351] bcm2835-clk 20101000.cprman: found enabled pll_div pllc_per - marking it as critical
> [    2.059427] bcm2835-clk 20101000.cprman: found enabled pll_div plld_core - marking it as critical
> [    2.068590] bcm2835-clk 20101000.cprman: found enabled pll_div plld_per - marking it as critical
> [    2.077724] bcm2835-clk 20101000.cprman: found enabled pll_div pllh_pix - marking it as critical

Yeah, pllh_pix isn't critical, though.  We want it to get turned off
when the driver asks to disable its clock, or we're going to just waste
a pile of power.

I'm sending out a patch that marks the VPU clock as critical (it's the
also AXI bus, so it certainly is critical), which should solve your
aux_uart clock disabling problem, I think.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 818 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-rpi-kernel/attachments/20160426/cd74e134/attachment.sig>