Skip to content

clk: samsung: zumapro: Add the PERIC0 USIs - #77

Closed
W4HO-ham wants to merge 10 commits into
derekross:omamobilefrom
W4HO-ham:tegu-peric0-usi
Closed

W4HO-ham wants to merge 10 commits into
derekross:omamobilefrom
W4HO-ham:tegu-peric0-usi

Conversation

@W4HO-ham

@W4HO-ham W4HO-ham commented Oct 27, 2026 •

Copy link
Copy Markdown
Contributor

Stacked on #76 (APM), #75, #74 and #73. Rebased onto omamobile after #40's merge. Only the last two commits are new: dt-bindings: clock: google,zumapro: Add the PERIC0 USIs and clk: samsung: zumapro: Add the PERIC0 USIs.

Upstream's CMU_PERIC0 describes only the GPIO path. This adds USI2 to USI6, the Pixel 9a's I2C buses:

USI Bus On the 9a
USI2 10890000.i2c (hsi2c_2) ultrawide camera
USI3 108a0000.i2c (hsi2c_3) front camera
USI4 108b0000.i2c (hsi2c_4) camera power chips, torch
USI5 108c0000.i2c (hsi2c_5) main camera
USI6 108d0000.i2c (hsi2c_6) fuel gauge, charger
  • Read-only state: each USI's user mux and divider are read-only, and its IPCLK, PCLK and RSTNSYNC gates are read-only gates, as is the PERIC0 sysreg gate. The GPIO gate and peric0_regs are unchanged.
  • CMU_TOP's PERIC0 IP branch, which feeds the USIs, gets a mux, a read-only gate and a divider, ids 49-51.
  • Binding: PERIC0 gets an optional third input, ip, so existing device trees with two clocks stay valid. The PERIC0 ids grow from 2 to 28, appended, and the existing two are unchanged.

I checked every offset, width and parent list against zuma's CAL. That covers the 26 PERIC0 registers, the 4-bit USI dividers, and the TOP PERIC0_IP mux, gate and divider with their parents (shared0/4, shared2/2, shared3/2, spare).

Boot: the palmtop's integration kernel already registers PERIC0 this way, on the common path. Its 28 PERIC0 entries match this commit's by id and offset, a scripted comparison with nothing on either side alone. Its CMU_TOP PERIC0_IP gate is read-only too. Running now: all five zumapro_*_usi{2..6}_ipclk at 399.36 MHz with the five I2C controllers above as consumers. The three cameras stream through libcamera, the camera power chips and torch work, and the fuel gauge and charger report.

checkpatch --strict is clean on both commits, clk-zumapro.o builds with W=1, and dt_binding_check passes for the binding.

@derekross

Copy link
Copy Markdown
Owner

The PERIC0 USIs match the CAL (the TOP IP branch at 0x10ec/0x2110/0x18e4, the five user muxes, dividers and fifteen gates, and the sysreg gate), and making ip optional keeps the existing node valid. Nothing blocking; it lands after #76. Two small things for the re-push if you like: the header's PERIC0 order (USI6, SYSREG, USI4, USI2, USI3, USI5) could simply run USI2 to USI6 then SYSREG while nothing uses it, and a uname -r plus blob id for the boot.

CMU_G3D clocks the Tensor G4's Mali GPU: two PLLs (the stacks' and the
L2's), user muxes for CMU_TOP's G3D NoC, global and switch clocks, and
the GPU's core, coregroup and stacks clocks. Describe it in the common
Tensor G4 clock binding with its four inputs, and add CMU_TOP's three
G3D branches (mux, gate and divider each) and the SHARED0 and SHARED1
halves and thirds those branches select at the end of the TOP block
and the 19 CMU_G3D identifiers in a new block, leaving the existing ones
unchanged.

Link: https://android.googlesource.com/kernel/google-modules/soc/gs/+/204078758737776959ecc85c527a00fd2cabb62f/drivers/soc/google/cal-if/zuma/cmucal-node.c
Signed-off-by: W4HO <87742609+W4HO-ham@users.noreply.github.com>
Register CMU_G3D, the Mali GPU's clock unit, on the common
exynos_arm64_register_cmu() path with its bus input, and CMU_TOP's three
G3D branches it takes its inputs from. Firmware sets G3D up and keeps
its domain on: the two PLLs (PLL0822x, no rate table), the muxes and the
dividers are read-only, and the GPU's core, coregroup and stacks gates
and the CMU's own gate are read-only gates, as are CMU_TOP's three G3D
gates. The STACKS and L2_GLB dividers have no ratio field in the CAL
(only busy and auto-gating bits), so they are 1:1 factors. The GPU's
rate changes go through the ACPM firmware, not this CMU.

Every offset, width and parent list was checked against zuma's cmucal-
sfr.c and cmucal-node.c. On the Pixel 9a the integration kernel
7.3.0-rc6-omamobile-tegu-s2mpg15-m1-g991042e2eff3-dirty, whose clk-
zumapro.c (blob ae4948e0dd60) has these G3D and CMU_TOP G3D tables,
registers the 30 G3D clocks and runs Panthor on the core, coregroup and
stacks gates (fout_g3d at 149.9 MHz, gout_g3d_core at 159.7 MHz).

CMU_TOP's G3D GLB and SWITCH muxes also select SHARED0 and SHARED1
divided by 2 and 3, which the driver didn't register, so a mux left
on one of them would be an orphan reading 0 Hz. Add the four fixed
factors (the CAL's PLL_SHARED0_D2 and _D3, PLL_SHARED1_D2 and _D3).
On the Pixel 9a the bootloader leaves the GLB mux on SHARED0_D2:
mout/gout_g3d_glb read 1065.984 MHz and dout_g3d_glb 532.992 MHz;
NOCD and SWITCH are on SHARED2 (798.72 MHz), dout_g3d_nocd 133.12
and dout_g3d_switch 266.24 MHz.

Link: https://android.googlesource.com/kernel/google-modules/soc/gs/+/204078758737776959ecc85c527a00fd2cabb62f/drivers/soc/google/cal-if/zuma/cmucal-sfr.c
Link: https://android.googlesource.com/kernel/google-modules/soc/gs/+/204078758737776959ecc85c527a00fd2cabb62f/drivers/soc/google/cal-if/zuma/cmucal-node.c
Signed-off-by: W4HO <87742609+W4HO-ham@users.noreply.github.com>
CMU_HSI1 clocks the HSI1 block's GPIO bank (the modem's control lines,
the haptics reset). Describe it in the common Tensor G4 clock binding
with its three inputs: the oscillator, CMU_TOP's HSI1 NoC clock and the
board's 400 MHz ALT fixed clock. Allow a power domain for it as for
HSI0, which firmware keeps on today. Add CMU_TOP's HSI1 branch at the
end of the TOP block and the 7 CMU_HSI1 identifiers in a new block,
leaving the existing ones unchanged.

Link: https://android.googlesource.com/kernel/google-modules/soc/gs/+/204078758737776959ecc85c527a00fd2cabb62f/drivers/soc/google/cal-if/zuma/cmucal-node.c
Signed-off-by: W4HO <87742609+W4HO-ham@users.noreply.github.com>
Register CMU_HSI1 on the common exynos_arm64_register_cmu() path with
its bus input, and CMU_TOP's HSI1 NoC branch. Firmware keeps the domain
powered and owns its setup: the NoC and ALT user muxes, the local NoC mux
and the ALT divider are read-only, and the ALT, GPIO and CMU gates are
read-only gates. Only the GPIO bank's clock path is described; the PCIe
channel 0 clocks are not. The ALT gate's parent is the board's 400 MHz
hsi1-alt fixed clock, found by name as HSI0's ALT and TCXO inputs are.

Every offset, width and parent list was checked against zuma's
cmucal-sfr.c and cmucal-node.c.

Booted on the Pixel 9a:
7.3.0-rc6-omamobile-tegu-s2mpg15-m1-g991042e2eff3-dirty, whose
clk-zumapro.c (blob c1db3a4a7736) carries these clocks under that
tree's own clock ids.

Link: https://android.googlesource.com/kernel/google-modules/soc/gs/+/204078758737776959ecc85c527a00fd2cabb62f/drivers/soc/google/cal-if/zuma/cmucal-sfr.c
Link: https://android.googlesource.com/kernel/google-modules/soc/gs/+/204078758737776959ecc85c527a00fd2cabb62f/drivers/soc/google/cal-if/zuma/cmucal-node.c
Signed-off-by: W4HO <87742609+W4HO-ham@users.noreply.github.com>
CMU_PERIC1 clocks the PERIC1 GPIO bank and USI9, the I2C bus of the
Pixel 9a's NFC controller and haptics driver. Describe it in the common
Tensor G4 clock binding with its three inputs: the oscillator and
CMU_TOP's PERIC1 NoC and IP clocks. Add CMU_TOP's two PERIC1 branches at
the end of the TOP block and the 8 CMU_PERIC1 identifiers in a new
block, leaving the existing ones unchanged.

Link: https://android.googlesource.com/kernel/google-modules/soc/gs/+/204078758737776959ecc85c527a00fd2cabb62f/drivers/soc/google/cal-if/zuma/cmucal-node.c
Signed-off-by: W4HO <87742609+W4HO-ham@users.noreply.github.com>
Register CMU_PERIC1 on the common exynos_arm64_register_cmu() path with
its bus input, and CMU_TOP's PERIC1 NoC and IP branches. Firmware sets
the block up: the NoC and USI9 user muxes and the USI9 divider are
read-only, and the GPIO, sysreg and USI9 gates are read-only gates. USI9
is the I2C bus of the Pixel 9a's NFC controller and haptics driver.

Every offset, width and parent list was checked against zuma's
cmucal-sfr.c and cmucal-node.c.

Booted on the Pixel 9a:
7.3.0-rc6-omamobile-tegu-s2mpg15-m1-g991042e2eff3-dirty, whose
clk-zumapro.c (blob c1db3a4a7736) carries these clocks under that
tree's own clock ids.

Link: https://android.googlesource.com/kernel/google-modules/soc/gs/+/204078758737776959ecc85c527a00fd2cabb62f/drivers/soc/google/cal-if/zuma/cmucal-sfr.c
Link: https://android.googlesource.com/kernel/google-modules/soc/gs/+/204078758737776959ecc85c527a00fd2cabb62f/drivers/soc/google/cal-if/zuma/cmucal-node.c
Signed-off-by: W4HO <87742609+W4HO-ham@users.noreply.github.com>
CMU_APM clocks the always-on GPIO banks, the APM mailbox and its
internal memory. Describe it in the common Tensor G4 clock binding with
the oscillator as its only input, as on gs101 and gs201, and add its 20
identifiers in a new block, leaving the existing ones unchanged.

Link: https://android.googlesource.com/kernel/google-modules/soc/gs/+/204078758737776959ecc85c527a00fd2cabb62f/drivers/soc/google/cal-if/zuma/cmucal-node.c
Signed-off-by: W4HO <87742609+W4HO-ham@users.noreply.github.com>
Register CMU_APM, which clocks the always-on GPIO banks, the APM mailbox
and its internal memory, on the common exynos_arm64_register_cmu() path.
Firmware sets it up and keeps it on: the FUNC and FUNCSRC muxes and the
NoC and USI1 UART dividers are read-only, and its gates are read-only
gates.

Its sources are the oscillator, from the DT, and the always-on ALV PLL's
four dividers, as fixed rates like gs101's. The CAL gives nominal figures
for the zuma family (787 MHz for the ALV PLL, 400, 200, 100 and 50 MHz
for its dividers, against a 26 MHz oscillator); the ALV PLL runs at 32
times the real 24.576 MHz oscillator, 786.432 MHz, so the dividers are
given as exact fractions of that. The external pad input is not
described, as on gs101.

Every offset, width and parent list was checked against zuma's
cmucal-sfr.c and cmucal-node.c.

Booted on the Pixel 9a:
7.3.0-rc6-omamobile-tegu-s2mpg15-m1-g991042e2eff3-dirty, whose
clk-zumapro.c (blob c1db3a4a7736) carries these clocks under that
tree's own clock ids.

Link: https://android.googlesource.com/kernel/google-modules/soc/gs/+/204078758737776959ecc85c527a00fd2cabb62f/drivers/soc/google/cal-if/zuma/cmucal-sfr.c
Link: https://android.googlesource.com/kernel/google-modules/soc/gs/+/204078758737776959ecc85c527a00fd2cabb62f/drivers/soc/google/cal-if/zuma/cmucal-node.c
Signed-off-by: W4HO <87742609+W4HO-ham@users.noreply.github.com>
CMU_PERIC0 also clocks USI2 to USI6: on the Pixel 9a the I2C buses of
the three cameras, the camera power chips and torch, and the battery
fuel gauge and charger. Their source is CMU_TOP's PERIC0 IP clock:
describe it as an optional third PERIC0 input, so existing device trees
with two stay valid, and add CMU_TOP's PERIC0 IP branch at the end of
the TOP block and the 26 new PERIC0 identifiers after the existing two.

Link: https://android.googlesource.com/kernel/google-modules/soc/gs/+/204078758737776959ecc85c527a00fd2cabb62f/drivers/soc/google/cal-if/zuma/cmucal-node.c
The PERIC0 ids run USI2 to USI6, then the sysreg gate.

Signed-off-by: W4HO <87742609+W4HO-ham@users.noreply.github.com>
Describe USI2 to USI6 in CMU_PERIC0, the Pixel 9a's I2C buses for the
three cameras, the camera power chips and torch, and the battery fuel
gauge and charger, and CMU_TOP's PERIC0 IP branch that feeds them.
Firmware sets them up: the USI user muxes and dividers are read-only,
and each USI's IPCLK, PCLK and RSTNSYNC gates and the PERIC0 sysreg gate
are read-only gates. The GPIO gate stays as it was.

Every offset, width and parent list was checked against zuma's
cmucal-sfr.c and cmucal-node.c.

Booted on the Pixel 9a:
7.3.0-rc6-omamobile-tegu-s2mpg15-m1-g991042e2eff3-dirty, whose
clk-zumapro.c (blob c1db3a4a7736) carries these clocks under that
tree's own clock ids.

Link: https://android.googlesource.com/kernel/google-modules/soc/gs/+/204078758737776959ecc85c527a00fd2cabb62f/drivers/soc/google/cal-if/zuma/cmucal-sfr.c
Link: https://android.googlesource.com/kernel/google-modules/soc/gs/+/204078758737776959ecc85c527a00fd2cabb62f/drivers/soc/google/cal-if/zuma/cmucal-node.c
Signed-off-by: W4HO <87742609+W4HO-ham@users.noreply.github.com>
@derekross

Copy link
Copy Markdown
Owner

The re-push looks right: USI2 to USI6, then SYSREG, and nr_clk_ids follows it. One small thing came with it. In the bindings commit (dt-bindings: clock: google,zumapro: Add the PERIC0 USIs), the new sentence "The PERIC0 ids run USI2 to USI6, then the sysreg gate." went in after the Link: line. Git now reads the Link: as body text, and git log -1 --format=%B <commit> | git interpret-trailers --parse lists only the Signed-off-by:. Could you move that sentence up into the paragraph above, or just drop it (the header shows the order), so the Link: sits right above your sign-off? Nothing else in the stack needs to change.

@derekross

Copy link
Copy Markdown
Owner

Landed through #79, which carried this commit with the rest of the stack. Thanks!

@derekross derekross closed this Oct 27, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants