rpi-4.7.y enable clock manager and modify all subsystems to use it #1628

clivem · 2016-08-31T20:10:12Z

Placeholder.... This is what I think we have to merge to move all of the subsystems over to using clock manager in rpi-4.7.y.

@msperl I think this is turning more into a "loose ends" to have clock manager enabled by default, and move all of the subsystems over to using it, rather than an "backport upstream clockmgr/I2S" thing. I'm still testing, but I'm not seeing any issues at the moment.

Undo the REVERT of the patches already reverted in downstream rpi-4.7.y to get us back to upstream. b3b4815 d4481e2 d4c84d9 f66b42f 658794c 01506b7 fb8d849 6019d04 b99821f 224a465 846ee11 Signed-off-by: DigitalDreamtime <[email protected]>

Enable the clock manager by default. Signed-off-by: Martin Sperl <[email protected]>

Removed clock registers from reg and added clock to i2s devicetree node. Signed-off-by: Martin Sperl <[email protected]>

Upstream mandates the use of "serial" as name inside the device-tree. Note that this does not affect the use of the alias - there uart0 and uart1 are permissible. This also fixed the patch references inside the overlays for the rpi3. Signed-off-by: Martin Sperl <[email protected]>

Moved uart0 to use new clock framework. At the same time removed the fixed clock clk_uart0 and the corresponding references, as the clock is now read directly from the configured clocks. If it ever becomes necessary the following can get added to modify the base clock rate later: assigned-clocks = <&clocks BCM2835_CLOCK_UART>; assigned-clock-rates = <X>; Signed-off-by: Martin Sperl <[email protected]>

use spi0 to use new clock framwork Signed-off-by: Martin Sperl <[email protected]>

move i2c to use the new clock framework instead of the fixed clock Signed-off-by: Martin Sperl <[email protected]>

Move PWM to use the new clock framework Signed-off-by: Martin Sperl <[email protected]>

Fix the auxiliar clock provider to use the clock framework Signed-off-by: Martin Sperl <[email protected]>

Fix the register range of the spi-aux devices, as they do not control the gate. Signed-off-by: Martin Sperl <[email protected]>

Move to new aux-uart driver (from upstream) and enable it correctly in the device tree Note that this may also solve the "serial not 115200 baud" issue on the rpi3... With an extension to the driver (requesting clock change notifications) the driver could change the baud divider when such a clock change occurs... It still would require some communication between the firmware and the kernel, so that this would get propagated. Signed-off-by: Martin Sperl <[email protected]>

Remove no longer necessary fixed pwm clock. Signed-off-by: Martin Sperl <[email protected]>

Use clock manager instead of self-made clockmanager. Also fix some error paths that showd up during development (especially missing release of dma resources on rmmod) Signed-off-by: Martin Sperl <[email protected]>

Move smi to use clock framework. Signed-off-by: Martin Sperl <[email protected]>

Move mmc to use clock framework. For some reason the emmc clock dirver does not claim the clock - enable count stays at 0 for the emmc clock. Signed-off-by: Martin Sperl <[email protected]>

Move the sdhost device to use the clock framework. Unfortunately I can not test it on my CM, as it seem as if there is no overlay that would enable sdhost and change the gpio-mux to the correct ALT0 (from Alt3 from mmc). Signed-off-by: Martin Sperl <[email protected]>

Remove clk_core from device tree as it is no longer used. Any "modification" to the core_clock in the bootloader is now automatically detected via the cnew clockmgr. Signed-off-by: Martin Sperl <[email protected]>

Revert the i2s-mmap dt overlay (7ee829f). Signed-off-by: DigitalDreamtime <[email protected]>

Moved uart0 to use new clock framework. At the same time removed the fixed clock clk_uart0 and the corresponding references, as the clock is now read directly from the configured clocks. If it ever becomes necessary the following can get added to modify the base clock rate later: assigned-clocks = <&clocks BCM2835_CLOCK_UART>; assigned-clock-rates = <X>; Signed-off-by: DigitalDreamtime <[email protected]>

Remove clk_core from device tree as it is no longer used. Any "modification" to the core_clock in the bootloader is now automatically detected via the cnew clockmgr. Signed-off-by: DigitalDreamtime <[email protected]>

msperl · 2016-08-31T20:49:08Z

On comment: imo the "bcm2835-i2s: REVERT the reverted patches" patch is slightly ugly - maybe just revert each of the reverts separately to document what actually happened instead of referencing in the comment.

But it is up to @popcornmix to decide what he prefers to merge...

Besides that I guess we should now be good enough to use the clock-framework for everything.
(Besides some intensive testing - especially SMI, as the corresponding patch is untested!)

This finally allows to use vc4 graphics together with i2s.

If this works fine in 4.7, then it should also be possible to cherry-pick these patches in 4.4, but I guess that may be seen as too risky.

msperl · 2016-09-01T07:16:16Z

here the quick branch that just did the reverts individually:
https://github.com/msperl/linux-rpi/commits/rpi-4.7.y-clock-only
This:

7064620 Revert "Revert "ASoC: bcm2835: Add S16_LE support via packed DMA transfers""
d53e90d Revert "Revert "ASoC: bcm2835: setup clock only if CPU is clock master""
02771a5 Revert "Revert "ASoC: bcm2835: add 24bit support""
bc5f27c Revert "Revert "ASoC: bcm2835: move to use the clock framework""
b79165f Revert "bcm2835-i2s: get base address for DMA from devicetree"
0abba71 Revert "bcm2835-i2s: add 24bit support, update bclk_ratio to more correct values"
c434816 Revert "bcm2835-i2s: setup clock only if CPU is clock master"
a23e7bf Revert "bcm2835-i2s: Eliminate debugfs directory error"
3f5e417 Revert "bcm2835-i2s: Register PCM device"
f7eabb0 Revert "bcm2835-i2s: Enable MMAP support via a DT property"
2341c37 Revert "bcm2835-i2s: Reduce the TX DREQ threshold"

only replaces this:
d0b57d0 bcm2835-i2s: REVERT the reverted patches

Note that:
git diff remotes/DigitalDreamtimeLtd/rpi-4.7.y-clockmgr3 remotes/origin/rpi-4.7.y-clock-only produces no output, so the branches are identical in content.

Note: I have not tested this yet - compile is running...

@popcornmix: the question is: do you want a pull request of this patchset?

clivem · 2016-09-01T12:58:34Z

@msperl Yes, I know. I was lazy with the "revert the reverts" patchset, bundling it into one commit. ;)
I figured as they rebase the non-current branches, ie. when 4.7.2->4.7.3, it would matter as the history would disappear, nothing would be reverted to start with, so wouldn't need to revert the reverts. ;)

@msperl I'm thinking I want to ask @popcornmix to also cherry-pick this 2b8dfa3 commit from rpi-4.4.y to rpi-4.7.y. What do you think? (I think any of the other Anholt clock commits are specific to his VC4 stuff and can wait until he rebases his 4.4 code to 4.7.)

popcornmix · 2016-09-01T13:59:56Z

At some point after merging the reverts and "reverts of reverts" will be dropped.
Individual reverts of each commit will make this slightly easy to pair up, so is preferable, but I imagine I'll deal with it either way.

msperl · 2016-09-01T14:36:28Z

@popcornmix : Created #1629 with reverted reverts instead of the "bulk revert".

Please choose which one you prefer to merge.

msperl · 2016-09-01T14:42:17Z

@clivem : the patch by Eric with regards to DSI/PLL reconfig probably needs to get in when all those DSI patches go in.

In principle I guess the next step would be for eric to update his VC4 patchset for 4.7, but I guess we should be able to cherry-pick the VC4 4.8 patches immediately:

git log --oneline v4.8-rc4 ^rpi-4.7.y-clock-only --no-merges  -- ./drivers/gpu/drm/vc4/
63fe9bb drm/vc4: Delete unnecessary checks before two function calls
a20d5fa drm/vc4: Fix a "the the" typo in a comment.
20e48fd drm/vc4: Fix definition of QPU_R_MS_REV_FLAGS
7363cee drm/vc4: Add a getparam to signal support for branches.
6d45c81 drm/vc4: Add support for branching in shader validation.
93aa9ae drm/vc4: Add a bitmap of branch targets during shader validation.
d0566c2 drm/vc4: Move validation's current/max ip into the validation struct.
af71379 drm/vc4: Add a getparam ioctl for getting the V3D identity regs.
deb4765 drm/vc4: remove redundant ret status check
1bf59f1 drm/vc4: Implement precise vblank timestamping.
7a10096 drm/vc4: Bind the HVS before we bind the individual CRTCs.
027a697 gpu: drm: vc4_hdmi: add missing of_node_put after calling of_parse_phandle
9397776 drm: vc4: enable XBGR8888 and ABGR8888 pixel formats
f510f34 drm/vc4: Remove unused connector
398e979 drm/vc4: Remove open-coded drm_connector_register_all()
b2d61d6 drm: vc4: Rely on the default ->best_encoder() behavior
5e84c26 drm/atomic-helper: Massage swap_state signature somewhat
7ea7728 drm/core: Change declaration for gamma_set.
a9402df drm/vc4: clean up error exit path on failed dpi_connector allocation
2f196b7 drm/atomic: Add drm_atomic_crtc_state_for_each_plane_state
833cd78 drm/vc4: Use for_each_plane_in_state
b82caaf drm/vc4: Use lockless gem BO free callback
db36972 drm/vc4: Use drm_gem_object_unreference_unlocked

I am not sure if there would be some need for some other (support) patches to make these really work...

msperl · 2016-09-01T15:31:45Z

a quick check (individual cherry pick + conflict check + compile check) shows that this one does not apply cleanly:

7ea7728 drm/core: Change declaration for gamma_set.
- this has the following prerequisites (to apply without issues):
  - 3abb263 drm/cirrus: Drop redundnant gamma size check
  - 6136bfb drm/atomic-helper: Fixup kerneldoc for drm_atomic_helper_legacy_gamma_set
  - 5d943aa drm: Consolidate crtc arrays in drm_atomic_state
  - b8b5342 drm: Consolidate plane arrays in drm_atomic_state
  - 63e83c1 drm: Consolidate connector arrays in drm_atomic_state
  - 2e7a570 drm/doc: Appease sphinx
  - be9174a drm/atomic-helper: use for_each_*_in_state more
  - a0909cc drm/atomic: Handle funcs->best_encoder == NULL case
  - f55f170 drm/atomic-helper: Do not call ->mode_fixup for CRTC which will be disabled
5e84c26 drm/atomic-helper: Massage swap_state signature somewhat
- 76c6dcc Merge branch 'virtio-gpu-for-airlied' of git://git.kraxel.org/linux into drm-next
  this is a merge and (probably) produces lots of conflicts...
- bbbed88 virtio-gpu: switch to atomic cursor interfaces
- e7cf096 virtio-gpu: add atomic_commit function
- a288c1e drm/virtio: use drm_crtc_send_vblank_event()

so for the vc4 stuff eric needs to look into it...

But anyway: this is outside the scope of this pull

clivem · 2016-09-01T16:02:02Z

so for the vc4 stuff eric needs to look into it...

Yes, I'd rather stay well away from anything to do with vc4. I am not going to be able to do any testing with all of my "headless" configurations....

clivem · 2016-09-01T16:02:49Z

Closing, in favour of #1629

DigitalDreamtime and others added 20 commits August 31, 2016 20:58

ARM: bcm2708: enable clock manager by default

76f119e

Enable the clock manager by default. Signed-off-by: Martin Sperl <[email protected]>

ARM: 2708: i2s: removed clock registers and added clock

abffab6

Removed clock registers from reg and added clock to i2s devicetree node. Signed-off-by: Martin Sperl <[email protected]>

ARM: bcm2708: move spi0 to use new clock framework

1d1f32d

use spi0 to use new clock framwork Signed-off-by: Martin Sperl <[email protected]>

ARM: bcm2708: move i2c to use new clock framework

2f565ef

move i2c to use the new clock framework instead of the fixed clock Signed-off-by: Martin Sperl <[email protected]>

ARM: bcm2708: move pwm to new clock framework

1c2f14f

Move PWM to use the new clock framework Signed-off-by: Martin Sperl <[email protected]>

ARM: bcm2708: fix auxiliar clock provider to use clock framework

a9a3cee

Fix the auxiliar clock provider to use the clock framework Signed-off-by: Martin Sperl <[email protected]>

ARM: bcm2708: fix register range of spi-aux

3abaa3d

Fix the register range of the spi-aux devices, as they do not control the gate. Signed-off-by: Martin Sperl <[email protected]>

ARM: bcm2708: remove pwm clock

d8fe04a

Remove no longer necessary fixed pwm clock. Signed-off-by: Martin Sperl <[email protected]>

MISC: bcm2835: smi: use clock manager and fix reload issues

7302cde

Use clock manager instead of self-made clockmanager. Also fix some error paths that showd up during development (especially missing release of dma resources on rmmod) Signed-off-by: Martin Sperl <[email protected]>

ARM: bcm2835: move smi to use clock framework

a524e40

Move smi to use clock framework. Signed-off-by: Martin Sperl <[email protected]>

ARM: bcm2708: move mmc to use clock framework

a5405ff

Move mmc to use clock framework. For some reason the emmc clock dirver does not claim the clock - enable count stays at 0 for the emmc clock. Signed-off-by: Martin Sperl <[email protected]>

ARM: bcm2708: remove clk_core as it is no longer used

fc7be1c

Remove clk_core from device tree as it is no longer used. Any "modification" to the core_clock in the bootloader is now automatically detected via the cnew clockmgr. Signed-off-by: Martin Sperl <[email protected]>

Revert "bcm2835-i2s: Enable MMAP support via a DT property"

1fb2daf

Revert the i2s-mmap dt overlay (7ee829f). Signed-off-by: DigitalDreamtime <[email protected]>

CM3: ARM: bcm2708: remove clk_core as it is no longer used

5fb9ffe

Remove clk_core from device tree as it is no longer used. Any "modification" to the core_clock in the bootloader is now automatically detected via the cnew clockmgr. Signed-off-by: DigitalDreamtime <[email protected]>

msperl mentioned this pull request Sep 1, 2016

rpi-4.7.y enable clock manager and modify all subsystems to use it - with reverts #1629

Merged

clivem closed this Sep 1, 2016

folkertvanheusden mentioned this pull request Jun 26, 2023

UVC gadget driver broken #5520

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

rpi-4.7.y enable clock manager and modify all subsystems to use it #1628

rpi-4.7.y enable clock manager and modify all subsystems to use it #1628

clivem commented Aug 31, 2016

msperl commented Aug 31, 2016

msperl commented Sep 1, 2016

clivem commented Sep 1, 2016 •

edited

Loading

popcornmix commented Sep 1, 2016

msperl commented Sep 1, 2016

msperl commented Sep 1, 2016

msperl commented Sep 1, 2016 •

edited

Loading

clivem commented Sep 1, 2016

clivem commented Sep 1, 2016 •

edited

Loading

rpi-4.7.y enable clock manager and modify all subsystems to use it #1628

rpi-4.7.y enable clock manager and modify all subsystems to use it #1628

Conversation

clivem commented Aug 31, 2016

msperl commented Aug 31, 2016

msperl commented Sep 1, 2016

clivem commented Sep 1, 2016 • edited Loading

popcornmix commented Sep 1, 2016

msperl commented Sep 1, 2016

msperl commented Sep 1, 2016

msperl commented Sep 1, 2016 • edited Loading

clivem commented Sep 1, 2016

clivem commented Sep 1, 2016 • edited Loading

clivem commented Sep 1, 2016 •

edited

Loading

msperl commented Sep 1, 2016 •

edited

Loading

clivem commented Sep 1, 2016 •

edited

Loading