On Tue, Sep 16, 2025 at 11:55:17AM -0400, Frank Li wrote:
On Mon, Sep 15, 2025 at 09:53:06PM +0200, Dario Binacchi wrote:
quoted
Support the touchscreen-glitch-threshold-ns property. Unlike the
generic description in touchscreen.yaml, this controller maps the
provided value to one of four discrete thresholds internally.
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
---
Changes in v3:
- Remove the final part of the description that refers to
implementation details.
.../bindings/input/touchscreen/fsl,imx6ul-tsc.yaml | 12 ++++++++++++
1 file changed, 12 insertions(+)
diff --git a/Documentation/devicetree/bindings/input/touchscreen/fsl,imx6ul-tsc.yaml b/Documentation/devicetree/bindings/input/touchscreen/fsl,imx6ul-tsc.yaml
index 678756ad0f92..6214d8be5a99 100644
--- a/Documentation/devicetree/bindings/input/touchscreen/fsl,imx6ul-tsc.yaml
+++ b/Documentation/devicetree/bindings/input/touchscreen/fsl,imx6ul-tsc.yaml
@@ -62,6 +62,18 @@ properties:
description: Number of data samples which are averaged for each read.
enum: [ 1, 4, 8, 16, 32 ]
+ touchscreen-glitch-threshold-ns:
+ description: |
+ Unlike the generic property defined in touchscreen.yaml, this
+ controller does not allow arbitrary values. Internally the value is
+ converted to IPG clock cycles and mapped to one of four discrete
+ thresholds exposed by the TSC_DEBUG_MODE2 register:
+
+ 0: 8191 IPG cycles
+ 1: 4095 IPG cycles
+ 2: 2047 IPG cycles
+ 3: 1023 IPG cycles
+
You have to use ns here. You can caculate in driver to match to closed one.
That is what he is saying it is doing - "internally the value is
converted to IPG clock cycles" and so on. Your repeated misunderstanding
of this though points out that maybe the description is still lacking?
Dario, how about:
| The gitch threshold in nanoseconds. Drivers must convert this value to
| IPG clock cycles and map it to one of the four discrete thresholds
| exposed by the TSC_DEBUG_MODE2 register:
|
| 0: 8191 IPG cycles
| 1: 4095 IPG cycles
| 2: 2047 IPG cycles
| 3: 1023 IPG cycles
?
I dropped the bit about arbitrary values, due to my comment on the other
version a few mins ago.