On Sat 01 May 16:09 CDT 2021, Marijn Suijten wrote:

Hi Bjorn,

On 4/29/21 11:15 PM, Bjorn Andersson wrote:

quoted

The Light Pulse Generator (LPG) is a PWM-block found in a wide range of
PMICs from Qualcomm. It can operate on fixed parameters or based on a
lookup-table, altering the duty cycle over time - which provides the
means for e.g. hardware assisted transitions of LED brightness.

Signed-off-by: Bjorn Andersson <redacted>
---

Changes since v6:
- Moved code into drivers/leds/rgb/
- Reverted to earlier qcom,dtest handling to support routing pwm signals
   through dtest lines.
- Remember the duration of each step of the pattern, rather than adding up and
   then dividing when the value is used.
- Added missing error prints on DT parse errors.
- Added sm8150[lb] and made led source and atc presence optional
- Added missing parenthesis around (len + 1) / 2 in search for hi_pause in the
   pattern.

  drivers/leds/Kconfig             |    3 +
  drivers/leds/Makefile            |    3 +
  drivers/leds/rgb/leds-qcom-lpg.c | 1286 ++++++++++++++++++++++++++++++
  3 files changed, 1292 insertions(+)
  create mode 100644 drivers/leds/rgb/leds-qcom-lpg.c

diff --git a/drivers/leds/Kconfig b/drivers/leds/Kconfig
index 49d99cb084db..8ab06b3f162d 100644
--- a/drivers/leds/Kconfig
+++ b/drivers/leds/Kconfig

@@ -933,6 +933,9 @@ source "drivers/leds/blink/Kconfig"
  comment "Flash and Torch LED drivers"
  source "drivers/leds/flash/Kconfig"
+comment "RGB LED drivers"
+source "drivers/leds/rgb/Kconfig"

It looks like this file is not included in any of the patches.

Sorry about that, seems like I missed adding them to the commit :(

quoted

+
  comment "LED Triggers"
  source "drivers/leds/trigger/Kconfig"

diff --git a/drivers/leds/Makefile b/drivers/leds/Makefile
index 7e604d3028c8..8cad0465aae0 100644
--- a/drivers/leds/Makefile
+++ b/drivers/leds/Makefile

@@ -106,6 +106,9 @@ obj-$(CONFIG_LEDS_USER)			+= uleds.o
  # Flash and Torch LED Drivers
  obj-$(CONFIG_LEDS_CLASS_FLASH)		+= flash/
+# RGB LED Drivers
+obj-$(CONFIG_LEDS_CLASS_MULTICOLOR)	+= rgb/

This file appears to be missing from this patch(set), too.

Right, missed both the new files.

quoted

+static int lpg_lut_store(struct lpg *lpg, struct led_pattern *pattern,
+			 size_t len, unsigned int *lo_idx, unsigned int *hi_idx)
+{
+	unsigned int idx;
+	u16 val;
+	int i;
+
+	/* Hardware does not behave when LO_IDX == HI_IDX */
+	if (len == 1)
+		return -EINVAL;
+
+	idx = bitmap_find_next_zero_area(lpg->lut_bitmap, lpg->lut_size,
+					 0, len, 0);
+	if (idx >= lpg->lut_size)
+		return -ENOMEM;
+
+	for (i = 0; i < len; i++) {
+		val = pattern[i].brightness;
+
+		regmap_bulk_write(lpg->map, lpg->lut_base + LPG_LUT_REG(idx + i), &val, 1);

This and the other regmap_bulk_write in lpg_apply_pwm_value used sizeof(val)
before.  As far as I'm aware qcom-spmi-pmic specifies 16-bit addresses
(.reg_bits) but 8-bit register sizes (.val_bits).  Writing one register
means only 8 out of 16 bits in u16 val are written?

You're right and I should have used test pattern that shown that I lost
that 9th bit...

I'll restore the sizeof()

quoted

+static void lpg_apply_lut_control(struct lpg_channel *chan)
+{
+	struct lpg *lpg = chan->lpg;
+	unsigned int hi_pause;
+	unsigned int lo_pause;
+	unsigned int step;
+	unsigned int conf = 0;
+	unsigned int lo_idx = chan->pattern_lo_idx;
+	unsigned int hi_idx = chan->pattern_hi_idx;
+	int pattern_len;
+
+	if (!chan->ramp_enabled || chan->pattern_lo_idx == chan->pattern_hi_idx)
+		return;
+
+	pattern_len = hi_idx - lo_idx + 1 > +
+	step = chan->ramp_tick_ms;

Since this is not dividing a full pattern duration by pattern_len anymore,
that variable is now never read and best removed.

Ok

quoted

+static int lpg_parse_channel(struct lpg *lpg, struct device_node *np,
+			     struct lpg_channel **channel)
+{
+	struct lpg_channel *chan;
+	u32 color = LED_COLOR_ID_GREEN;
+	u32 reg;
+	int ret;
+
+	ret = of_property_read_u32(np, "reg", &reg);
+	if (ret || !reg || reg > lpg->num_channels) {
+		dev_err(lpg->dev, "invalid reg of %pOFn\n", np);

Like \"color\" below, escape reg with \"reg\"?

Sounds good.

quoted

+static int lpg_add_led(struct lpg *lpg, struct device_node *np)
+{
+	struct led_classdev *cdev;
+	struct device_node *child;
+	struct mc_subled *info;
+	struct lpg_led *led;
+	const char *state;
+	int num_channels;
+	u32 color = 0;
+	int ret;
+	int i;
+
+	ret = of_property_read_u32(np, "color", &color);
+	if (ret < 0 && ret != -EINVAL) {
+		dev_err(lpg->dev, "failed to parse \"color\" of %pOF\n", np);
+		return ret;
+	}
+
+	if (color == LED_COLOR_ID_MULTI)

Since this driver now lives under rgb/, and is specifically for RGB leds
(afaik), should this and the rest of the code use LED_COLOR_ID_RGB instead?
There was a patch floating around on (if I remember correctly) ##linux-msm
by Luca Weiss that performs the conversion, with some related changes.

I thought MULTICOLOR was the right color, but reading the comment on the
defines it seems RGB is the color I actually want. Will check out
z3ntu's changes as well.

quoted

+static int lpg_init_lut(struct lpg *lpg)
+{
+	const struct lpg_data *data = lpg->data;
+	size_t bitmap_size;
+
+	if (!data->lut_base)
+		return 0;
+
+	lpg->lut_base = data->lut_base;
+	lpg->lut_size = data->lut_size;
+
+	bitmap_size = BITS_TO_BYTES(lpg->lut_size);
+	lpg->lut_bitmap = devm_kzalloc(lpg->dev, bitmap_size, GFP_KERNEL);
+	if (!lpg->lut_bitmap)
+		return -ENOMEM;
+
+	bitmap_clear(lpg->lut_bitmap, 0, lpg->lut_size);

devm_kzalloc already zeroes the bitmap.  Is it necessary to clear it again
(assuming a "cleared" bitmap is implementation-dependent and does not imply
zeroed memory) or could the memory be allocated with devm_kalloc instead?

You're right, I can drop the explicit clear of the bitmap.

Thanks,
Bjorn