[v4,0/4] Add device links to clocks
mbox series

Message ID 20190108161940.4814-1-miquel.raynal@bootlin.com
Headers show
Series
  • Add device links to clocks
Related show

Message

Miquel Raynal Jan. 8, 2019, 4:19 p.m. UTC
Hello,

While working on suspend to RAM feature, I ran into troubles multiple
times when clocks where not suspending/resuming at the desired time. I
had a look at the core and I think the same logic as in the
regulator's core may be applied here to (very easily) fix this issue:
using device links.

The only additional change I had to do was to always (when available)
populate the device entry of the core clock structure so that it could
be used later. This is the purpose of patch 1. Patch 2 actually adds
support for device links.

Here is a step-by-step explanation of how links are managed, following
Maxime Ripard's suggestion.


The order of probe has no importance because the framework already
handles orphaned clocks so let's be simple and say there are two root
clocks, not depending on anything, that are probed first: xtal0 and
xtal1. None of these clocks have a parent, there is no device link in
the game, yet.

   +----------------+            +----------------+
   |                |            |                |
   |                |            |                |
   |   xtal0 core   |            |   xtal1 core   |
   |                |            |                |
   |                |            |                |
   +-------^^-------+            +-------^^-------+
           ||                            ||
           ||                            ||
   +----------------+            +----------------+
   |                |            |                |
   |   xtal0 clk    |            |   xtal1 clk    |
   |                |            |                |
   +----------------+            +----------------+

Then, a peripheral clock periph0 is probed. His parent is xtal1. The
clock_register_*() call will run __clk_init_parent() and a link between
periph0's core and xtal1's core will be created and stored in
periph0's core->parent_clk_link entry.

   +----------------+            +----------------+
   |                |            |                |
   |                |            |                |
   |   xtal0 core   |            |   xtal1 core   |
   |                |            |                |
   |                |            |                |
   +-------^^-------+            +-------^^-------+
           ||                            ||
           ||                            ||
   +----------------+            +----------------+
   |                |            |                |
   |   xtal0 clk    |            |   xtal1 clk    |
   |                |            |                |
   +----------------+            +-------^--------+
                                         |
                                         |
                          +--------------+
                          |   ->parent_clk_link
                          |
                  +----------------+
                  |                |
                  |                |
                  |  periph0 core  |
                  |                |
                  |                |
                  +-------^^-------+
                          ||
                          ||
                  +----------------+
                  |                |
                  |  periph0 clk 0 |
                  |                |
                  +----------------+

Then, device0 is probed and "get" the periph0 clock. clk_get() will be
called and a struct clk will be instantiated for device0 (called in
the figure clk 1). A link between device0 and the new clk 1 instance of
periph0 will be created and stored in the clk->consumer_link entry.

   +----------------+            +----------------+
   |                |            |                |
   |                |            |                |
   |   xtal0 core   |            |   xtal1 core   |
   |                |            |                |
   |                |            |                |
   +-------^^-------+            +-------^^-------+
           ||                            ||
           ||                            ||
   +----------------+            +----------------+
   |                |            |                |
   |   xtal0 clk    |            |   xtal1 clk    |
   |                |            |                |
   +----------------+            +-------^--------+
                                         |
                                         |
                          +--------------+
                          |   ->parent_clk_link
                          |
                  +----------------+
                  |                |
                  |                |
                  |  periph0 core  |
                  |                <-------------+
                  |                <-------------|
                  +-------^^-------+            ||
                          ||                    ||
                          ||                    ||
                  +----------------+    +----------------+
                  |                |    |                |
                  |  periph0 clk 0 |    |  periph0 clk 1 |
                  |                |    |                |
                  +----------------+    +----------------+
                                                |
                                                | ->consumer_link
                                                |
                                                |
                                                |
                                        +-------v--------+
                                        |    device0     |
                                        +----------------+

Right now, device0 is linked to periph0, itself linked to xtal1 so
everything is fine.

Now let's get some fun: the new parent of periph0 is xtal1. The process
will call clk_reparent(), periph0's core->parent_clk_link will be
destroyed and a new link to xtal1 will be setup and stored. The
situation is now that device0 is linked to periph0 and periph0 is
linked to xtal1, so the dependency between device0 and xtal1 is still
clear.

   +----------------+            +----------------+
   |                |            |                |
   |                |            |                |
   |   xtal0 core   |            |   xtal1 core   |
   |                |            |                |
   |                |            |                |
   +-------^^-------+            +-------^^-------+
           ||                            ||
           ||                            ||
   +----------------+            +----------------+
   |                |            |                |
   |   xtal0 clk    |            |   xtal1 clk    |
   |                |            |                |
   +-------^--------+            +----------------+
           |
           |                           \ /
           +----------------------------x
      ->parent_clk_link   |            / \
                          |
                  +----------------+
                  |                |
                  |                |
                  |  periph0 core  |
                  |                <-------------+
                  |                <-------------|
                  +-------^^-------+            ||
                          ||                    ||
                          ||                    ||
                  +----------------+    +----------------+
                  |                |    |                |
                  |  periph0 clk 0 |    |  periph0 clk 1 |
                  |                |    |                |
                  +----------------+    +----------------+
                                                |
                                                | ->consumer_link
                                                |
                                                |
                                                |
                                        +-------v--------+
                                        |    device0     |
                                        +----------------+

I assume periph0 cannot be removed while there are devices using it,
same for xtal0.

What can happen is that device0 'put' the clock periph0. The relevant
link is deleted and the clk instance dropped.

   +----------------+            +----------------+
   |                |            |                |
   |                |            |                |
   |   xtal0 core   |            |   xtal1 core   |
   |                |            |                |
   |                |            |                |
   +-------^^-------+            +-------^^-------+
           ||                            ||
           ||                            ||
   +----------------+            +----------------+
   |                |            |                |
   |   xtal0 clk    |            |   xtal1 clk    |
   |                |            |                |
   +-------^--------+            +----------------+
           |
           |                           \ /
           +----------------------------x
      ->parent_clk_link   |            / \
                          |
                  +----------------+
                  |                |
                  |                |
                  |  periph0 core  |
                  |                |
                  |                |
                  +-------^^-------+
                          ||
                          ||
                  +----------------+
                  |                |
                  |  periph0 clk 0 |
                  |                |
                  +----------------+

Now we can unregister periph0: link with the parent will be destroyed
and the clock may be safely removed.

   +----------------+            +----------------+
   |                |            |                |
   |                |            |                |
   |   xtal0 core   |            |   xtal1 core   |
   |                |            |                |
   |                |            |                |
   +-------^^-------+            +-------^^-------+
           ||                            ||
           ||                            ||
   +----------------+            +----------------+
   |                |            |                |
   |   xtal0 clk    |            |   xtal1 clk    |
   |                |            |                |
   +----------------+            +----------------+


This is my understanding of the common clock framework and how links
can be added to it.

As a result, here are the links created during the boot of an
ESPRESSObin:

----->8-----
marvell-armada-3700-tbg-clock d0013200.tbg: Linked as a consumer to d0013800.pinctrl:xtal-clk
marvell-armada-3700-tbg-clock d0013200.tbg: Dropping the link to d0013800.pinctrl:xtal-clk
marvell-armada-3700-tbg-clock d0013200.tbg: Linked as a consumer to d0013800.pinctrl:xtal-clk
marvell-armada-3700-periph-clock d0013000.nb-periph-clk: Linked as a consumer to d0013200.tbg
marvell-armada-3700-periph-clock d0013000.nb-periph-clk: Linked as a consumer to d0013800.pinctrl:xtal-clk
marvell-armada-3700-periph-clock d0018000.sb-periph-clk: Linked as a consumer to d0013200.tbg
mvneta d0030000.ethernet: Linked as a consumer to d0018000.sb-periph-clk
xhci-hcd d0058000.usb: Linked as a consumer to d0018000.sb-periph-clk
xenon-sdhci d00d0000.sdhci: Linked as a consumer to d0013000.nb-periph-clk
xenon-sdhci d00d0000.sdhci: Dropping the link to d0013000.nb-periph-clk
mvebu-uart d0012000.serial: Linked as a consumer to d0013800.pinctrl:xtal-clk
advk-pcie d0070000.pcie: Linked as a consumer to d0018000.sb-periph-clk
xenon-sdhci d00d0000.sdhci: Linked as a consumer to d0013000.nb-periph-clk
xenon-sdhci d00d0000.sdhci: Linked as a consumer to regulator.1
cpu cpu0: Linked as a consumer to d0013000.nb-periph-clk
cpu cpu0: Dropping the link to d0013000.nb-periph-clk
cpu cpu0: Linked as a consumer to d0013000.nb-periph-clk
-----8<-----

Thanks,
Miquèl

Changes since v3:
=================
* Rebased on top of Stephen's 'clk-parent-rewrite' branch. Stephen
  already updated and took the patch 'clk: core: clarify the check for
  runtime PM' so it is not present in this series anymore.
* Updated the code to fit with the new core. Kept the helpers that
  were added to clk/clk.c (turning them static) for more readability.
* While working on clocks, I discovered a typo in an a3700-tbg driver
  error message. A patch has been added to correct this typo.

Changes since v2:
=================
* Fixed compilation issue when not using the common clock framework:
  removed the static keyword in front of clk_link/unlink_consumer()
  dummy definitions in clkdev.c.

Changes since v1:
=================
* Add clock->clock links, not only device->clock links.
* Helpers renamed:
  > clk_{link,unlink}_hierarchy()
  > clk_{link,unlink}_consumer()
* Add two patches to pass a "struct device" to the clock registration
  helper. This way device links may work between clocks themselves
  (otherwise the link is not created).


Miquel Raynal (4):
  clk: core: link consumer with clock driver
  clk: mvebu: armada-37xx-tbg: fix error message
  clk: mvebu: armada-37xx-tbg: fill the device entry when registering
    the clocks
  clk: mvebu: armada-37xx-xtal: fill the device entry when registering
    the clock

 drivers/clk/clk.c                    | 50 +++++++++++++++++++++++++++-
 drivers/clk/mvebu/armada-37xx-tbg.c  |  8 +++--
 drivers/clk/mvebu/armada-37xx-xtal.c |  3 +-
 3 files changed, 56 insertions(+), 5 deletions(-)

Comments

Stephen Boyd April 11, 2019, 11:34 p.m. UTC | #1
Quoting Miquel Raynal (2019-01-08 08:19:36)
> Hello,
> 
> While working on suspend to RAM feature, I ran into troubles multiple
> times when clocks where not suspending/resuming at the desired time. I
> had a look at the core and I think the same logic as in the
> regulator's core may be applied here to (very easily) fix this issue:
> using device links.
> 
> The only additional change I had to do was to always (when available)
> populate the device entry of the core clock structure so that it could
> be used later. This is the purpose of patch 1. Patch 2 actually adds
> support for device links.
> 
> Here is a step-by-step explanation of how links are managed, following
> Maxime Ripard's suggestion.
> 
> 
> The order of probe has no importance because the framework already
> handles orphaned clocks so let's be simple and say there are two root
> clocks, not depending on anything, that are probed first: xtal0 and
> xtal1. None of these clocks have a parent, there is no device link in
> the game, yet.
> 
>    +----------------+            +----------------+
>    |                |            |                |
>    |                |            |                |
>    |   xtal0 core   |            |   xtal1 core   |
>    |                |            |                |
>    |                |            |                |
>    +-------^^-------+            +-------^^-------+
>            ||                            ||
>            ||                            ||
>    +----------------+            +----------------+
>    |                |            |                |
>    |   xtal0 clk    |            |   xtal1 clk    |
>    |                |            |                |
>    +----------------+            +----------------+
> 
> Then, a peripheral clock periph0 is probed. His parent is xtal1. The
> clock_register_*() call will run __clk_init_parent() and a link between
> periph0's core and xtal1's core will be created and stored in
> periph0's core->parent_clk_link entry.
> 
>    +----------------+            +----------------+
>    |                |            |                |
>    |                |            |                |
>    |   xtal0 core   |            |   xtal1 core   |
>    |                |            |                |
>    |                |            |                |
>    +-------^^-------+            +-------^^-------+
>            ||                            ||
>            ||                            ||
>    +----------------+            +----------------+
>    |                |            |                |
>    |   xtal0 clk    |            |   xtal1 clk    |
>    |                |            |                |
>    +----------------+            +-------^--------+
>                                          |
>                                          |
>                           +--------------+
>                           |   ->parent_clk_link
>                           |
>                   +----------------+
>                   |                |
>                   |                |
>                   |  periph0 core  |
>                   |                |
>                   |                |
>                   +-------^^-------+
>                           ||
>                           ||
>                   +----------------+
>                   |                |
>                   |  periph0 clk 0 |
>                   |                |
>                   +----------------+
> 
> Then, device0 is probed and "get" the periph0 clock. clk_get() will be
> called and a struct clk will be instantiated for device0 (called in
> the figure clk 1). A link between device0 and the new clk 1 instance of
> periph0 will be created and stored in the clk->consumer_link entry.
> 
>    +----------------+            +----------------+
>    |                |            |                |
>    |                |            |                |
>    |   xtal0 core   |            |   xtal1 core   |
>    |                |            |                |
>    |                |            |                |
>    +-------^^-------+            +-------^^-------+
>            ||                            ||
>            ||                            ||
>    +----------------+            +----------------+
>    |                |            |                |
>    |   xtal0 clk    |            |   xtal1 clk    |
>    |                |            |                |
>    +----------------+            +-------^--------+
>                                          |
>                                          |
>                           +--------------+
>                           |   ->parent_clk_link
>                           |
>                   +----------------+
>                   |                |
>                   |                |
>                   |  periph0 core  |
>                   |                <-------------+
>                   |                <-------------|
>                   +-------^^-------+            ||
>                           ||                    ||
>                           ||                    ||
>                   +----------------+    +----------------+
>                   |                |    |                |
>                   |  periph0 clk 0 |    |  periph0 clk 1 |
>                   |                |    |                |
>                   +----------------+    +----------------+
>                                                 |
>                                                 | ->consumer_link
>                                                 |
>                                                 |
>                                                 |
>                                         +-------v--------+
>                                         |    device0     |
>                                         +----------------+
> 
> Right now, device0 is linked to periph0, itself linked to xtal1 so
> everything is fine.
> 
> Now let's get some fun: the new parent of periph0 is xtal1. The process
> will call clk_reparent(), periph0's core->parent_clk_link will be
> destroyed and a new link to xtal1 will be setup and stored. The
> situation is now that device0 is linked to periph0 and periph0 is
> linked to xtal1, so the dependency between device0 and xtal1 is still
> clear.
> 
>    +----------------+            +----------------+
>    |                |            |                |
>    |                |            |                |
>    |   xtal0 core   |            |   xtal1 core   |
>    |                |            |                |
>    |                |            |                |
>    +-------^^-------+            +-------^^-------+
>            ||                            ||
>            ||                            ||
>    +----------------+            +----------------+
>    |                |            |                |
>    |   xtal0 clk    |            |   xtal1 clk    |
>    |                |            |                |
>    +-------^--------+            +----------------+
>            |
>            |                           \ /
>            +----------------------------x
>       ->parent_clk_link   |            / \
>                           |
>                   +----------------+
>                   |                |
>                   |                |
>                   |  periph0 core  |
>                   |                <-------------+
>                   |                <-------------|
>                   +-------^^-------+            ||
>                           ||                    ||
>                           ||                    ||
>                   +----------------+    +----------------+
>                   |                |    |                |
>                   |  periph0 clk 0 |    |  periph0 clk 1 |
>                   |                |    |                |
>                   +----------------+    +----------------+
>                                                 |
>                                                 | ->consumer_link
>                                                 |
>                                                 |
>                                                 |
>                                         +-------v--------+
>                                         |    device0     |
>                                         +----------------+
> 
> I assume periph0 cannot be removed while there are devices using it,
> same for xtal0.
> 
> What can happen is that device0 'put' the clock periph0. The relevant
> link is deleted and the clk instance dropped.
> 
>    +----------------+            +----------------+
>    |                |            |                |
>    |                |            |                |
>    |   xtal0 core   |            |   xtal1 core   |
>    |                |            |                |
>    |                |            |                |
>    +-------^^-------+            +-------^^-------+
>            ||                            ||
>            ||                            ||
>    +----------------+            +----------------+
>    |                |            |                |
>    |   xtal0 clk    |            |   xtal1 clk    |
>    |                |            |                |
>    +-------^--------+            +----------------+
>            |
>            |                           \ /
>            +----------------------------x
>       ->parent_clk_link   |            / \
>                           |
>                   +----------------+
>                   |                |
>                   |                |
>                   |  periph0 core  |
>                   |                |
>                   |                |
>                   +-------^^-------+
>                           ||
>                           ||
>                   +----------------+
>                   |                |
>                   |  periph0 clk 0 |
>                   |                |
>                   +----------------+
> 
> Now we can unregister periph0: link with the parent will be destroyed
> and the clock may be safely removed.
> 
>    +----------------+            +----------------+
>    |                |            |                |
>    |                |            |                |
>    |   xtal0 core   |            |   xtal1 core   |
>    |                |            |                |
>    |                |            |                |
>    +-------^^-------+            +-------^^-------+
>            ||                            ||
>            ||                            ||
>    +----------------+            +----------------+
>    |                |            |                |
>    |   xtal0 clk    |            |   xtal1 clk    |
>    |                |            |                |
>    +----------------+            +----------------+
> 
> 
> This is my understanding of the common clock framework and how links
> can be added to it.
> 
> As a result, here are the links created during the boot of an
> ESPRESSObin:
> 

Sorry this patch series is taking way too long to get merged. It's
already mid-April!

So I still have some of the original questions I had from before, mostly
around circular parent chains between clk providers. For example, there
are clk providers that both provide clks to other providers and consume
clks from those providers. Does device links work gracefully here?

Just speaking from my own qcom experience, I can point to the PCIe PHY
that's a provider of a clk to GCC and a consumer of a clk in GCC. In
block diagram form this is:


      PCIE PHY                        GCC
   +--------------+          +-------------------------+
   |              |          |                         |
   |     PHY clk ->----------+---- gcc_pipe_clk ---+   |
   |              |          |                     |   |
   |              |          |                     |   |
   | pci_pipe_clk <----------|---------------------+   |
   |              |          |                         |
   +--------------+          +-------------------------+

The end result is that the PCIe PHY is a clk controller that provides
the PHY clk to GCC's gcc_pipe_clk and then it gets the same clk signal
back from GCC and uses it on the PCIe PHY's pci_pipe_clk input.

So is this is a problem?
Miquel Raynal May 21, 2019, 9:46 a.m. UTC | #2
Hi Stephen,

Stephen Boyd <sboyd@kernel.org> wrote on Thu, 11 Apr 2019 16:34:16
-0700:

> Quoting Miquel Raynal (2019-01-08 08:19:36)
> > Hello,
> > 
> > While working on suspend to RAM feature, I ran into troubles multiple
> > times when clocks where not suspending/resuming at the desired time. I
> > had a look at the core and I think the same logic as in the
> > regulator's core may be applied here to (very easily) fix this issue:
> > using device links.
> > 
> > The only additional change I had to do was to always (when available)
> > populate the device entry of the core clock structure so that it could
> > be used later. This is the purpose of patch 1. Patch 2 actually adds
> > support for device links.
> > 
> > Here is a step-by-step explanation of how links are managed, following
> > Maxime Ripard's suggestion.
> > 
> > 
> > The order of probe has no importance because the framework already
> > handles orphaned clocks so let's be simple and say there are two root
> > clocks, not depending on anything, that are probed first: xtal0 and
> > xtal1. None of these clocks have a parent, there is no device link in
> > the game, yet.
> > 
> >    +----------------+            +----------------+
> >    |                |            |                |
> >    |                |            |                |
> >    |   xtal0 core   |            |   xtal1 core   |
> >    |                |            |                |
> >    |                |            |                |
> >    +-------^^-------+            +-------^^-------+
> >            ||                            ||
> >            ||                            ||
> >    +----------------+            +----------------+
> >    |                |            |                |
> >    |   xtal0 clk    |            |   xtal1 clk    |
> >    |                |            |                |
> >    +----------------+            +----------------+
> > 
> > Then, a peripheral clock periph0 is probed. His parent is xtal1. The
> > clock_register_*() call will run __clk_init_parent() and a link between
> > periph0's core and xtal1's core will be created and stored in
> > periph0's core->parent_clk_link entry.
> > 
> >    +----------------+            +----------------+
> >    |                |            |                |
> >    |                |            |                |
> >    |   xtal0 core   |            |   xtal1 core   |
> >    |                |            |                |
> >    |                |            |                |
> >    +-------^^-------+            +-------^^-------+
> >            ||                            ||
> >            ||                            ||
> >    +----------------+            +----------------+
> >    |                |            |                |
> >    |   xtal0 clk    |            |   xtal1 clk    |
> >    |                |            |                |
> >    +----------------+            +-------^--------+
> >                                          |
> >                                          |
> >                           +--------------+
> >                           |   ->parent_clk_link
> >                           |
> >                   +----------------+
> >                   |                |
> >                   |                |
> >                   |  periph0 core  |
> >                   |                |
> >                   |                |
> >                   +-------^^-------+
> >                           ||
> >                           ||
> >                   +----------------+
> >                   |                |
> >                   |  periph0 clk 0 |
> >                   |                |
> >                   +----------------+
> > 
> > Then, device0 is probed and "get" the periph0 clock. clk_get() will be
> > called and a struct clk will be instantiated for device0 (called in
> > the figure clk 1). A link between device0 and the new clk 1 instance of
> > periph0 will be created and stored in the clk->consumer_link entry.
> > 
> >    +----------------+            +----------------+
> >    |                |            |                |
> >    |                |            |                |
> >    |   xtal0 core   |            |   xtal1 core   |
> >    |                |            |                |
> >    |                |            |                |
> >    +-------^^-------+            +-------^^-------+
> >            ||                            ||
> >            ||                            ||
> >    +----------------+            +----------------+
> >    |                |            |                |
> >    |   xtal0 clk    |            |   xtal1 clk    |
> >    |                |            |                |
> >    +----------------+            +-------^--------+
> >                                          |
> >                                          |
> >                           +--------------+
> >                           |   ->parent_clk_link
> >                           |
> >                   +----------------+
> >                   |                |
> >                   |                |
> >                   |  periph0 core  |
> >                   |                <-------------+
> >                   |                <-------------|
> >                   +-------^^-------+            ||
> >                           ||                    ||
> >                           ||                    ||
> >                   +----------------+    +----------------+
> >                   |                |    |                |
> >                   |  periph0 clk 0 |    |  periph0 clk 1 |
> >                   |                |    |                |
> >                   +----------------+    +----------------+
> >                                                 |
> >                                                 | ->consumer_link
> >                                                 |
> >                                                 |
> >                                                 |
> >                                         +-------v--------+
> >                                         |    device0     |
> >                                         +----------------+
> > 
> > Right now, device0 is linked to periph0, itself linked to xtal1 so
> > everything is fine.
> > 
> > Now let's get some fun: the new parent of periph0 is xtal1. The process
> > will call clk_reparent(), periph0's core->parent_clk_link will be
> > destroyed and a new link to xtal1 will be setup and stored. The
> > situation is now that device0 is linked to periph0 and periph0 is
> > linked to xtal1, so the dependency between device0 and xtal1 is still
> > clear.
> > 
> >    +----------------+            +----------------+
> >    |                |            |                |
> >    |                |            |                |
> >    |   xtal0 core   |            |   xtal1 core   |
> >    |                |            |                |
> >    |                |            |                |
> >    +-------^^-------+            +-------^^-------+
> >            ||                            ||
> >            ||                            ||
> >    +----------------+            +----------------+
> >    |                |            |                |
> >    |   xtal0 clk    |            |   xtal1 clk    |
> >    |                |            |                |
> >    +-------^--------+            +----------------+
> >            |
> >            |                           \ /
> >            +----------------------------x  
> >       ->parent_clk_link   |            / \  
> >                           |
> >                   +----------------+
> >                   |                |
> >                   |                |
> >                   |  periph0 core  |
> >                   |                <-------------+
> >                   |                <-------------|
> >                   +-------^^-------+            ||
> >                           ||                    ||
> >                           ||                    ||
> >                   +----------------+    +----------------+
> >                   |                |    |                |
> >                   |  periph0 clk 0 |    |  periph0 clk 1 |
> >                   |                |    |                |
> >                   +----------------+    +----------------+
> >                                                 |
> >                                                 | ->consumer_link
> >                                                 |
> >                                                 |
> >                                                 |
> >                                         +-------v--------+
> >                                         |    device0     |
> >                                         +----------------+
> > 
> > I assume periph0 cannot be removed while there are devices using it,
> > same for xtal0.
> > 
> > What can happen is that device0 'put' the clock periph0. The relevant
> > link is deleted and the clk instance dropped.
> > 
> >    +----------------+            +----------------+
> >    |                |            |                |
> >    |                |            |                |
> >    |   xtal0 core   |            |   xtal1 core   |
> >    |                |            |                |
> >    |                |            |                |
> >    +-------^^-------+            +-------^^-------+
> >            ||                            ||
> >            ||                            ||
> >    +----------------+            +----------------+
> >    |                |            |                |
> >    |   xtal0 clk    |            |   xtal1 clk    |
> >    |                |            |                |
> >    +-------^--------+            +----------------+
> >            |
> >            |                           \ /
> >            +----------------------------x  
> >       ->parent_clk_link   |            / \  
> >                           |
> >                   +----------------+
> >                   |                |
> >                   |                |
> >                   |  periph0 core  |
> >                   |                |
> >                   |                |
> >                   +-------^^-------+
> >                           ||
> >                           ||
> >                   +----------------+
> >                   |                |
> >                   |  periph0 clk 0 |
> >                   |                |
> >                   +----------------+
> > 
> > Now we can unregister periph0: link with the parent will be destroyed
> > and the clock may be safely removed.
> > 
> >    +----------------+            +----------------+
> >    |                |            |                |
> >    |                |            |                |
> >    |   xtal0 core   |            |   xtal1 core   |
> >    |                |            |                |
> >    |                |            |                |
> >    +-------^^-------+            +-------^^-------+
> >            ||                            ||
> >            ||                            ||
> >    +----------------+            +----------------+
> >    |                |            |                |
> >    |   xtal0 clk    |            |   xtal1 clk    |
> >    |                |            |                |
> >    +----------------+            +----------------+
> > 
> > 
> > This is my understanding of the common clock framework and how links
> > can be added to it.
> > 
> > As a result, here are the links created during the boot of an
> > ESPRESSObin:
> >   
> 
> Sorry this patch series is taking way too long to get merged. It's
> already mid-April!
> 
> So I still have some of the original questions I had from before, mostly
> around circular parent chains between clk providers. For example, there
> are clk providers that both provide clks to other providers and consume
> clks from those providers. Does device links work gracefully here?
> 
> Just speaking from my own qcom experience, I can point to the PCIe PHY
> that's a provider of a clk to GCC and a consumer of a clk in GCC. In
> block diagram form this is:
> 
> 
>       PCIE PHY                        GCC
>    +--------------+          +-------------------------+
>    |              |          |                         |
>    |     PHY clk ->----------+---- gcc_pipe_clk ---+   |
>    |              |          |                     |   |
>    |              |          |                     |   |
>    | pci_pipe_clk <----------|---------------------+   |
>    |              |          |                         |
>    +--------------+          +-------------------------+
> 
> The end result is that the PCIe PHY is a clk controller that provides
> the PHY clk to GCC's gcc_pipe_clk and then it gets the same clk signal
> back from GCC and uses it on the PCIe PHY's pci_pipe_clk input.
> 
> So is this is a problem?
> 

It's now my turn to get back on this topic.

I just put my noise back into this and for what I understand of the
clk subsystem, I think the situation you describe could be pictured
like this:


         +---------------+
         |               |
         |               |
         | PCIe PHY      |
         |               |
         |               |
         +-----^^--------+
               ||
               ||
         +---------------+
         |               |
         | pcie_pipe_clk |
         |               |
         +------^--------+
                |
                | ->parent_clk_link
                |
                |
         +---------------+
         |               |
         |               |
         | GCC           |
         |               |
         |               |
         +------^^-------+
                ||
                ||
         +---------------+
         |               |
         | gcc_pipe_clk  |
         |               |
         +------^--------+
                |
                | ->parent_clk_link
                |
                |
         +---------------+
         |               |
         |               |
         | PCIe PHY      |
         |               |
         |               |
         +------^^-------+
                ||
                ||
         +---------------+
         |               |
         | phy_clk       |
         |               |
         +---------------+


IMHO the fact that the first and third blocks are the same does not
interfere with device links.

Honestly, I cannot be 100% sure it won't break on qcom designs, maybe
the best would be to have someone to test. I don't have the relevant
hardware. Do you? It would be really helpful!

There is an entire PCIe series blocked, waiting for these device links
to be merged so it would help a lot if someone could test.

Thank you very much,
Miquèl
Miquel Raynal June 17, 2019, 9:57 a.m. UTC | #3
Hi Stephen,

Miquel Raynal <miquel.raynal@bootlin.com> wrote on Tue, 21 May 2019
11:46:44 +0200:

> Hi Stephen,
> 
> Stephen Boyd <sboyd@kernel.org> wrote on Thu, 11 Apr 2019 16:34:16
> -0700:
> 
> > Quoting Miquel Raynal (2019-01-08 08:19:36)  
> > > Hello,
> > > 
> > > While working on suspend to RAM feature, I ran into troubles multiple
> > > times when clocks where not suspending/resuming at the desired time. I
> > > had a look at the core and I think the same logic as in the
> > > regulator's core may be applied here to (very easily) fix this issue:
> > > using device links.
> > > 
> > > The only additional change I had to do was to always (when available)
> > > populate the device entry of the core clock structure so that it could
> > > be used later. This is the purpose of patch 1. Patch 2 actually adds
> > > support for device links.
> > > 
> > > Here is a step-by-step explanation of how links are managed, following
> > > Maxime Ripard's suggestion.
> > > 
> > > 
> > > The order of probe has no importance because the framework already
> > > handles orphaned clocks so let's be simple and say there are two root
> > > clocks, not depending on anything, that are probed first: xtal0 and
> > > xtal1. None of these clocks have a parent, there is no device link in
> > > the game, yet.
> > > 
> > >    +----------------+            +----------------+
> > >    |                |            |                |
> > >    |                |            |                |
> > >    |   xtal0 core   |            |   xtal1 core   |
> > >    |                |            |                |
> > >    |                |            |                |
> > >    +-------^^-------+            +-------^^-------+
> > >            ||                            ||
> > >            ||                            ||
> > >    +----------------+            +----------------+
> > >    |                |            |                |
> > >    |   xtal0 clk    |            |   xtal1 clk    |
> > >    |                |            |                |
> > >    +----------------+            +----------------+
> > > 
> > > Then, a peripheral clock periph0 is probed. His parent is xtal1. The
> > > clock_register_*() call will run __clk_init_parent() and a link between
> > > periph0's core and xtal1's core will be created and stored in
> > > periph0's core->parent_clk_link entry.
> > > 
> > >    +----------------+            +----------------+
> > >    |                |            |                |
> > >    |                |            |                |
> > >    |   xtal0 core   |            |   xtal1 core   |
> > >    |                |            |                |
> > >    |                |            |                |
> > >    +-------^^-------+            +-------^^-------+
> > >            ||                            ||
> > >            ||                            ||
> > >    +----------------+            +----------------+
> > >    |                |            |                |
> > >    |   xtal0 clk    |            |   xtal1 clk    |
> > >    |                |            |                |
> > >    +----------------+            +-------^--------+
> > >                                          |
> > >                                          |
> > >                           +--------------+
> > >                           |   ->parent_clk_link
> > >                           |
> > >                   +----------------+
> > >                   |                |
> > >                   |                |
> > >                   |  periph0 core  |
> > >                   |                |
> > >                   |                |
> > >                   +-------^^-------+
> > >                           ||
> > >                           ||
> > >                   +----------------+
> > >                   |                |
> > >                   |  periph0 clk 0 |
> > >                   |                |
> > >                   +----------------+
> > > 
> > > Then, device0 is probed and "get" the periph0 clock. clk_get() will be
> > > called and a struct clk will be instantiated for device0 (called in
> > > the figure clk 1). A link between device0 and the new clk 1 instance of
> > > periph0 will be created and stored in the clk->consumer_link entry.
> > > 
> > >    +----------------+            +----------------+
> > >    |                |            |                |
> > >    |                |            |                |
> > >    |   xtal0 core   |            |   xtal1 core   |
> > >    |                |            |                |
> > >    |                |            |                |
> > >    +-------^^-------+            +-------^^-------+
> > >            ||                            ||
> > >            ||                            ||
> > >    +----------------+            +----------------+
> > >    |                |            |                |
> > >    |   xtal0 clk    |            |   xtal1 clk    |
> > >    |                |            |                |
> > >    +----------------+            +-------^--------+
> > >                                          |
> > >                                          |
> > >                           +--------------+
> > >                           |   ->parent_clk_link
> > >                           |
> > >                   +----------------+
> > >                   |                |
> > >                   |                |
> > >                   |  periph0 core  |
> > >                   |                <-------------+
> > >                   |                <-------------|
> > >                   +-------^^-------+            ||
> > >                           ||                    ||
> > >                           ||                    ||
> > >                   +----------------+    +----------------+
> > >                   |                |    |                |
> > >                   |  periph0 clk 0 |    |  periph0 clk 1 |
> > >                   |                |    |                |
> > >                   +----------------+    +----------------+
> > >                                                 |
> > >                                                 | ->consumer_link
> > >                                                 |
> > >                                                 |
> > >                                                 |
> > >                                         +-------v--------+
> > >                                         |    device0     |
> > >                                         +----------------+
> > > 
> > > Right now, device0 is linked to periph0, itself linked to xtal1 so
> > > everything is fine.
> > > 
> > > Now let's get some fun: the new parent of periph0 is xtal1. The process
> > > will call clk_reparent(), periph0's core->parent_clk_link will be
> > > destroyed and a new link to xtal1 will be setup and stored. The
> > > situation is now that device0 is linked to periph0 and periph0 is
> > > linked to xtal1, so the dependency between device0 and xtal1 is still
> > > clear.
> > > 
> > >    +----------------+            +----------------+
> > >    |                |            |                |
> > >    |                |            |                |
> > >    |   xtal0 core   |            |   xtal1 core   |
> > >    |                |            |                |
> > >    |                |            |                |
> > >    +-------^^-------+            +-------^^-------+
> > >            ||                            ||
> > >            ||                            ||
> > >    +----------------+            +----------------+
> > >    |                |            |                |
> > >    |   xtal0 clk    |            |   xtal1 clk    |
> > >    |                |            |                |
> > >    +-------^--------+            +----------------+
> > >            |
> > >            |                           \ /
> > >            +----------------------------x    
> > >       ->parent_clk_link   |            / \    
> > >                           |
> > >                   +----------------+
> > >                   |                |
> > >                   |                |
> > >                   |  periph0 core  |
> > >                   |                <-------------+
> > >                   |                <-------------|
> > >                   +-------^^-------+            ||
> > >                           ||                    ||
> > >                           ||                    ||
> > >                   +----------------+    +----------------+
> > >                   |                |    |                |
> > >                   |  periph0 clk 0 |    |  periph0 clk 1 |
> > >                   |                |    |                |
> > >                   +----------------+    +----------------+
> > >                                                 |
> > >                                                 | ->consumer_link
> > >                                                 |
> > >                                                 |
> > >                                                 |
> > >                                         +-------v--------+
> > >                                         |    device0     |
> > >                                         +----------------+
> > > 
> > > I assume periph0 cannot be removed while there are devices using it,
> > > same for xtal0.
> > > 
> > > What can happen is that device0 'put' the clock periph0. The relevant
> > > link is deleted and the clk instance dropped.
> > > 
> > >    +----------------+            +----------------+
> > >    |                |            |                |
> > >    |                |            |                |
> > >    |   xtal0 core   |            |   xtal1 core   |
> > >    |                |            |                |
> > >    |                |            |                |
> > >    +-------^^-------+            +-------^^-------+
> > >            ||                            ||
> > >            ||                            ||
> > >    +----------------+            +----------------+
> > >    |                |            |                |
> > >    |   xtal0 clk    |            |   xtal1 clk    |
> > >    |                |            |                |
> > >    +-------^--------+            +----------------+
> > >            |
> > >            |                           \ /
> > >            +----------------------------x    
> > >       ->parent_clk_link   |            / \    
> > >                           |
> > >                   +----------------+
> > >                   |                |
> > >                   |                |
> > >                   |  periph0 core  |
> > >                   |                |
> > >                   |                |
> > >                   +-------^^-------+
> > >                           ||
> > >                           ||
> > >                   +----------------+
> > >                   |                |
> > >                   |  periph0 clk 0 |
> > >                   |                |
> > >                   +----------------+
> > > 
> > > Now we can unregister periph0: link with the parent will be destroyed
> > > and the clock may be safely removed.
> > > 
> > >    +----------------+            +----------------+
> > >    |                |            |                |
> > >    |                |            |                |
> > >    |   xtal0 core   |            |   xtal1 core   |
> > >    |                |            |                |
> > >    |                |            |                |
> > >    +-------^^-------+            +-------^^-------+
> > >            ||                            ||
> > >            ||                            ||
> > >    +----------------+            +----------------+
> > >    |                |            |                |
> > >    |   xtal0 clk    |            |   xtal1 clk    |
> > >    |                |            |                |
> > >    +----------------+            +----------------+
> > > 
> > > 
> > > This is my understanding of the common clock framework and how links
> > > can be added to it.
> > > 
> > > As a result, here are the links created during the boot of an
> > > ESPRESSObin:
> > >     
> > 
> > Sorry this patch series is taking way too long to get merged. It's
> > already mid-April!
> > 
> > So I still have some of the original questions I had from before, mostly
> > around circular parent chains between clk providers. For example, there
> > are clk providers that both provide clks to other providers and consume
> > clks from those providers. Does device links work gracefully here?
> > 
> > Just speaking from my own qcom experience, I can point to the PCIe PHY
> > that's a provider of a clk to GCC and a consumer of a clk in GCC. In
> > block diagram form this is:
> > 
> > 
> >       PCIE PHY                        GCC
> >    +--------------+          +-------------------------+
> >    |              |          |                         |
> >    |     PHY clk ->----------+---- gcc_pipe_clk ---+   |
> >    |              |          |                     |   |
> >    |              |          |                     |   |
> >    | pci_pipe_clk <----------|---------------------+   |
> >    |              |          |                         |
> >    +--------------+          +-------------------------+
> > 
> > The end result is that the PCIe PHY is a clk controller that provides
> > the PHY clk to GCC's gcc_pipe_clk and then it gets the same clk signal
> > back from GCC and uses it on the PCIe PHY's pci_pipe_clk input.
> > 
> > So is this is a problem?
> >   
> 
> It's now my turn to get back on this topic.
> 
> I just put my noise back into this and for what I understand of the
> clk subsystem, I think the situation you describe could be pictured
> like this:
> 
> 
>          +---------------+
>          |               |
>          |               |
>          | PCIe PHY      |
>          |               |
>          |               |
>          +-----^^--------+
>                ||
>                ||
>          +---------------+
>          |               |
>          | pcie_pipe_clk |
>          |               |
>          +------^--------+
>                 |
>                 | ->parent_clk_link
>                 |
>                 |
>          +---------------+
>          |               |
>          |               |
>          | GCC           |
>          |               |
>          |               |
>          +------^^-------+
>                 ||
>                 ||
>          +---------------+
>          |               |
>          | gcc_pipe_clk  |
>          |               |
>          +------^--------+
>                 |
>                 | ->parent_clk_link
>                 |
>                 |
>          +---------------+
>          |               |
>          |               |
>          | PCIe PHY      |
>          |               |
>          |               |
>          +------^^-------+
>                 ||
>                 ||
>          +---------------+
>          |               |
>          | phy_clk       |
>          |               |
>          +---------------+
> 
> 
> IMHO the fact that the first and third blocks are the same does not
> interfere with device links.
> 
> Honestly, I cannot be 100% sure it won't break on qcom designs, maybe
> the best would be to have someone to test. I don't have the relevant
> hardware. Do you? It would be really helpful!
> 
> There is an entire PCIe series blocked, waiting for these device links
> to be merged so it would help a lot if someone could test.
> 

Could you share the status of this series? Will it be applied for the
next merge window? I would really like to see this moving forward.

> Thank you very much,
> Miquèl


Thanks,
Miquèl
Miquel Raynal July 27, 2019, 8:53 a.m. UTC | #4
Hi Stephen,

Miquel Raynal <miquel.raynal@bootlin.com> wrote on Mon, 17 Jun 2019
11:57:03 +0200:

> Hi Stephen,
> 
> Miquel Raynal <miquel.raynal@bootlin.com> wrote on Tue, 21 May 2019
> 11:46:44 +0200:
> 
> > Hi Stephen,
> > 
> > Stephen Boyd <sboyd@kernel.org> wrote on Thu, 11 Apr 2019 16:34:16
> > -0700:
> >   
> > > Quoting Miquel Raynal (2019-01-08 08:19:36)    
> > > > Hello,
> > > > 
> > > > While working on suspend to RAM feature, I ran into troubles multiple
> > > > times when clocks where not suspending/resuming at the desired time. I
> > > > had a look at the core and I think the same logic as in the
> > > > regulator's core may be applied here to (very easily) fix this issue:
> > > > using device links.
> > > > 
> > > > The only additional change I had to do was to always (when available)
> > > > populate the device entry of the core clock structure so that it could
> > > > be used later. This is the purpose of patch 1. Patch 2 actually adds
> > > > support for device links.
> > > > 
> > > > Here is a step-by-step explanation of how links are managed, following
> > > > Maxime Ripard's suggestion.
> > > > 
> > > > 
> > > > The order of probe has no importance because the framework already
> > > > handles orphaned clocks so let's be simple and say there are two root
> > > > clocks, not depending on anything, that are probed first: xtal0 and
> > > > xtal1. None of these clocks have a parent, there is no device link in
> > > > the game, yet.
> > > > 
> > > >    +----------------+            +----------------+
> > > >    |                |            |                |
> > > >    |                |            |                |
> > > >    |   xtal0 core   |            |   xtal1 core   |
> > > >    |                |            |                |
> > > >    |                |            |                |
> > > >    +-------^^-------+            +-------^^-------+
> > > >            ||                            ||
> > > >            ||                            ||
> > > >    +----------------+            +----------------+
> > > >    |                |            |                |
> > > >    |   xtal0 clk    |            |   xtal1 clk    |
> > > >    |                |            |                |
> > > >    +----------------+            +----------------+
> > > > 
> > > > Then, a peripheral clock periph0 is probed. His parent is xtal1. The
> > > > clock_register_*() call will run __clk_init_parent() and a link between
> > > > periph0's core and xtal1's core will be created and stored in
> > > > periph0's core->parent_clk_link entry.
> > > > 
> > > >    +----------------+            +----------------+
> > > >    |                |            |                |
> > > >    |                |            |                |
> > > >    |   xtal0 core   |            |   xtal1 core   |
> > > >    |                |            |                |
> > > >    |                |            |                |
> > > >    +-------^^-------+            +-------^^-------+
> > > >            ||                            ||
> > > >            ||                            ||
> > > >    +----------------+            +----------------+
> > > >    |                |            |                |
> > > >    |   xtal0 clk    |            |   xtal1 clk    |
> > > >    |                |            |                |
> > > >    +----------------+            +-------^--------+
> > > >                                          |
> > > >                                          |
> > > >                           +--------------+
> > > >                           |   ->parent_clk_link
> > > >                           |
> > > >                   +----------------+
> > > >                   |                |
> > > >                   |                |
> > > >                   |  periph0 core  |
> > > >                   |                |
> > > >                   |                |
> > > >                   +-------^^-------+
> > > >                           ||
> > > >                           ||
> > > >                   +----------------+
> > > >                   |                |
> > > >                   |  periph0 clk 0 |
> > > >                   |                |
> > > >                   +----------------+
> > > > 
> > > > Then, device0 is probed and "get" the periph0 clock. clk_get() will be
> > > > called and a struct clk will be instantiated for device0 (called in
> > > > the figure clk 1). A link between device0 and the new clk 1 instance of
> > > > periph0 will be created and stored in the clk->consumer_link entry.
> > > > 
> > > >    +----------------+            +----------------+
> > > >    |                |            |                |
> > > >    |                |            |                |
> > > >    |   xtal0 core   |            |   xtal1 core   |
> > > >    |                |            |                |
> > > >    |                |            |                |
> > > >    +-------^^-------+            +-------^^-------+
> > > >            ||                            ||
> > > >            ||                            ||
> > > >    +----------------+            +----------------+
> > > >    |                |            |                |
> > > >    |   xtal0 clk    |            |   xtal1 clk    |
> > > >    |                |            |                |
> > > >    +----------------+            +-------^--------+
> > > >                                          |
> > > >                                          |
> > > >                           +--------------+
> > > >                           |   ->parent_clk_link
> > > >                           |
> > > >                   +----------------+
> > > >                   |                |
> > > >                   |                |
> > > >                   |  periph0 core  |
> > > >                   |                <-------------+
> > > >                   |                <-------------|
> > > >                   +-------^^-------+            ||
> > > >                           ||                    ||
> > > >                           ||                    ||
> > > >                   +----------------+    +----------------+
> > > >                   |                |    |                |
> > > >                   |  periph0 clk 0 |    |  periph0 clk 1 |
> > > >                   |                |    |                |
> > > >                   +----------------+    +----------------+
> > > >                                                 |
> > > >                                                 | ->consumer_link
> > > >                                                 |
> > > >                                                 |
> > > >                                                 |
> > > >                                         +-------v--------+
> > > >                                         |    device0     |
> > > >                                         +----------------+
> > > > 
> > > > Right now, device0 is linked to periph0, itself linked to xtal1 so
> > > > everything is fine.
> > > > 
> > > > Now let's get some fun: the new parent of periph0 is xtal1. The process
> > > > will call clk_reparent(), periph0's core->parent_clk_link will be
> > > > destroyed and a new link to xtal1 will be setup and stored. The
> > > > situation is now that device0 is linked to periph0 and periph0 is
> > > > linked to xtal1, so the dependency between device0 and xtal1 is still
> > > > clear.
> > > > 
> > > >    +----------------+            +----------------+
> > > >    |                |            |                |
> > > >    |                |            |                |
> > > >    |   xtal0 core   |            |   xtal1 core   |
> > > >    |                |            |                |
> > > >    |                |            |                |
> > > >    +-------^^-------+            +-------^^-------+
> > > >            ||                            ||
> > > >            ||                            ||
> > > >    +----------------+            +----------------+
> > > >    |                |            |                |
> > > >    |   xtal0 clk    |            |   xtal1 clk    |
> > > >    |                |            |                |
> > > >    +-------^--------+            +----------------+
> > > >            |
> > > >            |                           \ /
> > > >            +----------------------------x      
> > > >       ->parent_clk_link   |            / \      
> > > >                           |
> > > >                   +----------------+
> > > >                   |                |
> > > >                   |                |
> > > >                   |  periph0 core  |
> > > >                   |                <-------------+
> > > >                   |                <-------------|
> > > >                   +-------^^-------+            ||
> > > >                           ||                    ||
> > > >                           ||                    ||
> > > >                   +----------------+    +----------------+
> > > >                   |                |    |                |
> > > >                   |  periph0 clk 0 |    |  periph0 clk 1 |
> > > >                   |                |    |                |
> > > >                   +----------------+    +----------------+
> > > >                                                 |
> > > >                                                 | ->consumer_link
> > > >                                                 |
> > > >                                                 |
> > > >                                                 |
> > > >                                         +-------v--------+
> > > >                                         |    device0     |
> > > >                                         +----------------+
> > > > 
> > > > I assume periph0 cannot be removed while there are devices using it,
> > > > same for xtal0.
> > > > 
> > > > What can happen is that device0 'put' the clock periph0. The relevant
> > > > link is deleted and the clk instance dropped.
> > > > 
> > > >    +----------------+            +----------------+
> > > >    |                |            |                |
> > > >    |                |            |                |
> > > >    |   xtal0 core   |            |   xtal1 core   |
> > > >    |                |            |                |
> > > >    |                |            |                |
> > > >    +-------^^-------+            +-------^^-------+
> > > >            ||                            ||
> > > >            ||                            ||
> > > >    +----------------+            +----------------+
> > > >    |                |            |                |
> > > >    |   xtal0 clk    |            |   xtal1 clk    |
> > > >    |                |            |                |
> > > >    +-------^--------+            +----------------+
> > > >            |
> > > >            |                           \ /
> > > >            +----------------------------x      
> > > >       ->parent_clk_link   |            / \      
> > > >                           |
> > > >                   +----------------+
> > > >                   |                |
> > > >                   |                |
> > > >                   |  periph0 core  |
> > > >                   |                |
> > > >                   |                |
> > > >                   +-------^^-------+
> > > >                           ||
> > > >                           ||
> > > >                   +----------------+
> > > >                   |                |
> > > >                   |  periph0 clk 0 |
> > > >                   |                |
> > > >                   +----------------+
> > > > 
> > > > Now we can unregister periph0: link with the parent will be destroyed
> > > > and the clock may be safely removed.
> > > > 
> > > >    +----------------+            +----------------+
> > > >    |                |            |                |
> > > >    |                |            |                |
> > > >    |   xtal0 core   |            |   xtal1 core   |
> > > >    |                |            |                |
> > > >    |                |            |                |
> > > >    +-------^^-------+            +-------^^-------+
> > > >            ||                            ||
> > > >            ||                            ||
> > > >    +----------------+            +----------------+
> > > >    |                |            |                |
> > > >    |   xtal0 clk    |            |   xtal1 clk    |
> > > >    |                |            |                |
> > > >    +----------------+            +----------------+
> > > > 
> > > > 
> > > > This is my understanding of the common clock framework and how links
> > > > can be added to it.
> > > > 
> > > > As a result, here are the links created during the boot of an
> > > > ESPRESSObin:
> > > >       
> > > 
> > > Sorry this patch series is taking way too long to get merged. It's
> > > already mid-April!
> > > 
> > > So I still have some of the original questions I had from before, mostly
> > > around circular parent chains between clk providers. For example, there
> > > are clk providers that both provide clks to other providers and consume
> > > clks from those providers. Does device links work gracefully here?
> > > 
> > > Just speaking from my own qcom experience, I can point to the PCIe PHY
> > > that's a provider of a clk to GCC and a consumer of a clk in GCC. In
> > > block diagram form this is:
> > > 
> > > 
> > >       PCIE PHY                        GCC
> > >    +--------------+          +-------------------------+
> > >    |              |          |                         |
> > >    |     PHY clk ->----------+---- gcc_pipe_clk ---+   |
> > >    |              |          |                     |   |
> > >    |              |          |                     |   |
> > >    | pci_pipe_clk <----------|---------------------+   |
> > >    |              |          |                         |
> > >    +--------------+          +-------------------------+
> > > 
> > > The end result is that the PCIe PHY is a clk controller that provides
> > > the PHY clk to GCC's gcc_pipe_clk and then it gets the same clk signal
> > > back from GCC and uses it on the PCIe PHY's pci_pipe_clk input.
> > > 
> > > So is this is a problem?
> > >     
> > 
> > It's now my turn to get back on this topic.
> > 
> > I just put my noise back into this and for what I understand of the
> > clk subsystem, I think the situation you describe could be pictured
> > like this:
> > 
> > 
> >          +---------------+
> >          |               |
> >          |               |
> >          | PCIe PHY      |
> >          |               |
> >          |               |
> >          +-----^^--------+
> >                ||
> >                ||
> >          +---------------+
> >          |               |
> >          | pcie_pipe_clk |
> >          |               |
> >          +------^--------+
> >                 |
> >                 | ->parent_clk_link
> >                 |
> >                 |
> >          +---------------+
> >          |               |
> >          |               |
> >          | GCC           |
> >          |               |
> >          |               |
> >          +------^^-------+
> >                 ||
> >                 ||
> >          +---------------+
> >          |               |
> >          | gcc_pipe_clk  |
> >          |               |
> >          +------^--------+
> >                 |
> >                 | ->parent_clk_link
> >                 |
> >                 |
> >          +---------------+
> >          |               |
> >          |               |
> >          | PCIe PHY      |
> >          |               |
> >          |               |
> >          +------^^-------+
> >                 ||
> >                 ||
> >          +---------------+
> >          |               |
> >          | phy_clk       |
> >          |               |
> >          +---------------+
> > 
> > 
> > IMHO the fact that the first and third blocks are the same does not
> > interfere with device links.
> > 
> > Honestly, I cannot be 100% sure it won't break on qcom designs, maybe
> > the best would be to have someone to test. I don't have the relevant
> > hardware. Do you? It would be really helpful!
> > 
> > There is an entire PCIe series blocked, waiting for these device links
> > to be merged so it would help a lot if someone could test.
> >   
> 
> Could you share the status of this series? Will it be applied for the
> next merge window? I would really like to see this moving forward.

I know this series might have side effects despite the consequent
amount of time spent to write and test it, but I also think the
clk subsystem would really benefit from such change and handling
suspend to RAM support would be greatly enhanced. You seemed
interested at first and now not anymore, could I know why? I got
inspired by the regulators subsystem. It is not an idea of mine
that device links should be bring to clocks. Regulators are almost
as used as clocks so I really understand your fears but why not
applying this to -next very early during the -rc cycles and see
what happens? You'll have plenty of time to ask me to fix things
or even drop it off.

Kind regards,
Miquèl
Stephen Boyd Aug. 14, 2019, 6:41 p.m. UTC | #5
Quoting Miquel Raynal (2019-07-27 01:53:30)
> 
> I know this series might have side effects despite the consequent
> amount of time spent to write and test it, but I also think the
> clk subsystem would really benefit from such change and handling
> suspend to RAM support would be greatly enhanced. You seemed
> interested at first and now not anymore, could I know why? I got
> inspired by the regulators subsystem. It is not an idea of mine
> that device links should be bring to clocks. Regulators are almost
> as used as clocks so I really understand your fears but why not
> applying this to -next very early during the -rc cycles and see
> what happens? You'll have plenty of time to ask me to fix things
> or even drop it off.
> 

Ok, I'm back on this topic. Let me look at the latest code and see how
it works on a qcom platform I have in hand. If the device links look OK
then it should be good. I also want to make sure we're not holding a
nested pile of locks when we're adding the device links so that we don't
get some weird lockdep problems.