graphics/input: Add getevent input event monitor tool by HongChao6 · Pull Request #3434 · apache/nuttx-apps

HongChao6 · 2026-03-25T02:17:55Z

Summary

Add getevent, a command-line input event monitor that captures and displays real-time events from input devices.

What

getevent captures and displays real-time input events from mouse, touchscreen, and keyboard devices. It reads from input device nodes (/dev/mouse*, /dev/input*, /dev/kbd*) using non-blocking I/O with poll(), and prints structured event data via syslog.

Supported event types:

Mouse: button state, x/y coordinates, wheel (if enabled)
Touchscreen: multi-touch points with coordinates, flags, timestamps, and optional pressure/size detail
Keyboard: key code and event type

When / Why

This tool is useful during:

Driver bring-up: verify that a newly ported input driver produces correct events
Input debugging: inspect raw event data to diagnose issues like missed touches, incorrect coordinates, or stuck keys
System integration: confirm that input devices are properly registered and accessible from userspace

Without getevent, developers must write ad-hoc test programs or read device nodes manually to verify input behavior. This tool provides a ready-made, consistent way to do so.

Usage

# Auto-detect all input devices in /dev
getevent

# Monitor a specific mouse device
getevent -m /dev/mouse0

# Monitor a specific touchscreen
getevent -t /dev/input0

# Monitor a specific keyboard
getevent -k /dev/kbd0

# Combine multiple devices
getevent -m /dev/mouse0 -t /dev/input0 -k /dev/kbd0

Press Ctrl+C to stop.

Configuration

CONFIG_GRAPHICS_INPUT_GETEVENT=y            # Enable getevent
CONFIG_GRAPHICS_INPUT_GETEVENT_DETAIL_INFO=y # Optional: show touch pressure/size

Impact

Is new feature added? Is existing feature changed? YES - New getevent command-line tool added under graphics/input/.
Impact on user (will user need to adapt to change)? NO - This is a new optional tool, no existing functionality is changed.
Impact on build (will build process change)? NO - Only builds when CONFIG_GRAPHICS_INPUT_GETEVENT is explicitly enabled.
Impact on hardware (will arch(s) / board(s) / driver(s) change)? NO - No driver or architecture changes, reads existing input device nodes only.
Impact on documentation (is update required / provided)? YES - Documentation PR submitted: Documentation: Add getevent input event monitor documentation. nuttx#18615
Impact on security (any sort of implications)? NO
Impact on compatibility (backward/forward/interoperability)? NO - Purely additive, no existing code is modified (except adding getevent entry to Make.defs and Makefile).

Testing

I confirm that changes are verified on local setup and works as intended:

Build Host(s): Ubuntu 22.04.5 LTS, x86_64, GCC 13.4.0
Target(s): arm

Testing logs (runtime):

[getevent]: opening /dev/input0
[getevent]: ioctl TSIOC_GETMAXPOINTS: 1
[getevent]: opening /dev/kbd0
[getevent]: opening /dev/mouse0

[getevent]: touch event: /dev/input0
[getevent]:    npoints : 1
[getevent]: Point      : 0
[getevent]:      flags : 11
[getevent]:          x : 493
[getevent]:          y : 312
[getevent]:  timestamp : 31272800

[getevent]: keyboard event: /dev/kbd0
[getevent]:          type : 0 
[getevent]:          code : 57
[getevent]: keyboard event: /dev/kbd0
[getevent]:          type : 1
[getevent]:          code : 57

[getevent]: mouse event: /dev/mouse0
[getevent]:    buttons : 00
[getevent]:          x : 14
[getevent]:          y : -16

PR verification Self-Check

This PR introduces only one functional change.
I have updated all required description fields above.
My PR adheres to Contributing Guidelines and Documentation.
My PR is still work in progress (not ready for review).
My PR is ready for review and can be safely merged into a codebase.

cederom

Thank you @HongChao6 very useful utility :-)

Please read https://github.com/apache/nuttx/blob/master/CONTRIBUTING.md and update git commit descriptions :-)

I am wondering if nuttx-apps/testing/input would not be a better fit for this tool? The main purpose is testing / input not really graphics right? Also exiting code in graphics/input? What do you think folks? :-)

graphics/input/getevent/README.md

HongChao6 · 2026-03-25T06:34:25Z

Thank you @HongChao6 very useful utility :-)

Please read https://github.com/apache/nuttx/blob/master/CONTRIBUTING.md and update git commit descriptions :-)

I am wondering if nuttx-apps/testing/input would not be a better fit for this tool? The main purpose is testing / input not really graphics right? Also exiting code in graphics/input? What do you think folks? :-)

Thanks for the suggestion. The getevent tool shares the same input subsystem infrastructure with monkey and generator under graphics/input/. Moving getevent alone would split the input toolset.

If the community agrees that graphics/input is not the right place, I'd prefer to create a separate PR to move the entire graphics/input/ directory to testing/input/ together, rather than splitting them up in this PR.

What do you think?

graphics/input/Kconfig

acassis

@HongChao6 please create a Documentation about this Graphics Input Getevent, I think we also need some basic Documentation about Graphics Input in general

cederom · 2026-03-25T15:58:52Z

@acassis is this input tightly coupled with graphics? if not and can be used as general input then maybe we should create a separate and dedicated nuttx-apps/input ? :-)

acassis · 2026-03-25T16:07:03Z

@acassis is this input tightly coupled with graphics? if not and can be used as general input then maybe we should create a separate and dedicated nuttx-apps/input ? :-)

AFAIK Graphics Input is mostly used to create Monkey testing events, but a generic Input device tester

cederom · 2026-03-25T20:40:56Z

Okay not to delay here we can put where existing input resides :-) Just an idea to uncover this functionality into its own space :-)

cederom

TODO :-)

PR description.
Documentation.
Kconfig description.

cederom · 2026-03-25T21:23:15Z

I have created a dedicated issue for discussion of input location #3436 :-)

simbit18 · 2026-03-27T09:50:22Z

Hi @HongChao6, this error is not related to this PR

You will need to wait for this PR #3440 to be merged, after which you will need to rebase

====================================================================================
Configuration/Tool: sim/citest
2026-03-27 02:48:55
------------------------------------------------------------------------------------
  Cleaning...
  Configuring...
  Building NuttX...
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed

  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed

  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0

  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0

100 1646k  100 1646k    0     0  3078k      0 --:--:-- --:--:-- --:--:-- 5899k

100  143k    0  143k    0     0   286k      0 --:--:-- --:--:-- --:--:--  286k
  Running NuttX...
++ pwd
+ olddir=/github/workspace/sources/nuttx
+ nuttdir=/github/workspace/sources/nuttx/boards/sim/sim/sim/configs/citest/../../../../../../
+ cd /github/workspace/sources/nuttx/boards/sim/sim/sim/configs/citest/../../../../../../
+ confpath=/github/workspace/sources/nuttx/boards/sim/sim/sim/configs/citest/config.yaml
+ jsonconf=/github/workspace/sources/nuttx/boards/sim/sim/sim/configs/citest/session.json
+ testpath=/github/workspace/nuttx-ntfc/external/nuttx-testing
+ ntfc test --testpath=/github/workspace/nuttx-ntfc/external/nuttx-testing --confpath=/github/workspace/sources/nuttx/boards/sim/sim/sim/configs/citest/config.yaml --jsonconf=/github/workspace/sources/nuttx/boards/sim/sim/sim/configs/citest/session.json
/github/workspace/sources/nuttx/../nuttx/boards/sim/sim/sim/configs/citest/run.sh: line 35: ntfc: command not found
+ ret=127
+ echo 127
+ artifacts=/github/workspace/buildartifacts/sim/citest//ntfc
+ mkdir -p /github/workspace/buildartifacts/sim/citest//ntfc
127
+ rm -f pytest.debug.log
+ mv result /github/workspace/buildartifacts/sim/citest//ntfc
mv: cannot stat 'result': No such file or directory
+ cd /github/workspace/sources/nuttx
+ exit 127
  [1/1] Normalize sim/citest
====================================================================================

HongChao6 · 2026-03-27T10:56:25Z

Hi @HongChao6, this error is not related to this PR

You will need to wait for this PR #3440 to be merged, after which you will need to rebase

Thanks for the heads up! I'll rebase once #3440 is merged.

simbit18 · 2026-03-27T13:45:50Z

@HongChao6 please rebase

HongChao6 · 2026-03-30T09:04:28Z

@simbit18 The xtensa-03 failure is unrelated to this PR. It's caused by the Xtensa common-source integration update (nuttx#18614) which requires new Kconfig options for ESP32-S3 defconfigs. See esp_irq.c:125 compile error.

simbit18 · 2026-03-30T10:30:31Z

Hi @JianyuWang0623, could you please check?

====================================================================================
Configuration/Tool: lckfb-szpi-esp32s3/uvc
2026-03-30 08:01:51
------------------------------------------------------------------------------------
  Cleaning...
  Configuring...
  Building NuttX...
Error: ./common/espressif/esp_irq.c:125:4: error: #error "CONFIG_ARCH_MINIMAL_VECTORTABLE_DYNAMIC must be enabled for           Xtensa-based Espressif SoCs. Additionally, enable           CONFIG_ARCH_IRQ_TO_NDX and set CONFIG_ARCH_NUSER_INTERRUPTS to 2.          Run 'make menuconfig' to select and set these options."
  125 | #  error "CONFIG_ARCH_MINIMAL_VECTORTABLE_DYNAMIC must be enabled for \
      |    ^~~~~
ERROR: xtensa-esp32s3-elf-gcc failed: 1

Linux (xtensa-03)

JianyuWang0623 · 2026-03-30T12:08:35Z

Hi @JianyuWang0623, could you please check?

====================================================================================
Configuration/Tool: lckfb-szpi-esp32s3/uvc
2026-03-30 08:01:51
------------------------------------------------------------------------------------
  Cleaning...
  Configuring...
  Building NuttX...
Error: ./common/espressif/esp_irq.c:125:4: error: #error "CONFIG_ARCH_MINIMAL_VECTORTABLE_DYNAMIC must be enabled for           Xtensa-based Espressif SoCs. Additionally, enable           CONFIG_ARCH_IRQ_TO_NDX and set CONFIG_ARCH_NUSER_INTERRUPTS to 2.          Run 'make menuconfig' to select and set these options."
  125 | #  error "CONFIG_ARCH_MINIMAL_VECTORTABLE_DYNAMIC must be enabled for \
      |    ^~~~~
ERROR: xtensa-esp32s3-elf-gcc failed: 1

Linux (xtensa-03)

@simbit18 The fix commit is currently undergoing CI checks: apache/nuttx#18650

simbit18 · 2026-03-30T12:25:47Z

@JianyuWang0623 Thank you very much

linguini1 · 2026-03-30T15:18:49Z

Hi @HongChao6 , can you please include the impact and test sections in your PR from the contribution guide?

simbit18 · 2026-03-31T10:48:18Z

Hi @HongChao6 please rebase

PR apache/nuttx#18650 has been merged

graphics/input/getevent/getevent.c

linguini1 · 2026-03-31T12:47:22Z

Looks good besides a few small comments!

HongChao6 · 2026-04-01T02:02:05Z

Thanks for the review @linguini1! I agree with the GETEVENT_ERR macro and printf for show_usage — those changes make sense and I'll apply them.

Regarding GETEVENT_INFO and LOG_INFO: since getevent is an interactive command-line tool where the event data output is directly requested by the user, I'm wondering if printf (stdout) would be a better fit here as well, rather than syslog at any level. My reasoning:

Event data is user-facing output, not system log messages
Using printf avoids dependency on CONFIG_SYSLOG_LEVEL — with LOG_INFO, output could be silently filtered on systems with a higher default syslog threshold
This is consistent with how Android's getevent works (stdout)
syslog would still be used for actual errors via GETEVENT_ERR
What do you think? Happy to go with LOG_INFO if you prefer keeping syslog for everything.

linguini1 · 2026-04-01T02:14:19Z

Go with printf!

linguini1 · 2026-04-01T21:30:07Z

I suppose similarly, wouldn't you want to use fprintf to stderr instead of syslog errors? You can also remove the macros entirely and directly write printf since they're just wrappers.

HongChao6 · 2026-04-02T02:25:26Z

I suppose similarly, wouldn't you want to use fprintf to stderr instead of syslog errors? You can also remove the macros entirely and directly write printf since they're just wrappers.

Good point, I'll switch GETEVENT_ERR to use fprintf(stderr, ...) instead of syslog.

I'd prefer to keep the macros though — they provide a consistent [getevent]: prefix across all output, so if the format ever needs to change, it's a single-line edit. Without them, the prefix string would be duplicated in every call site.

I'll update and push shortly.

linguini1 · 2026-04-02T03:21:34Z

Forgot about the prefix, makes sense!

Add getevent utility for monitoring input events including mouse clicks/movement, multi-touch coordinates/pressure, and keyboard key presses. Supports automatic device detection in /dev and command-line device path specification with non-blocking I/O. Signed-off-by: liuhongchao <liuhongchao@xiaomi.com>

github-actions bot added Size: L Area: Graphics labels Mar 25, 2026

cederom requested changes Mar 25, 2026

View reviewed changes

cederom reviewed Mar 25, 2026

View reviewed changes

graphics/input/getevent/README.md Outdated Show resolved Hide resolved

cederom requested review from acassis, hartmannathan, linguini1, lupyuen, michallenc, raiden00pl, simbit18 and tmedicci March 25, 2026 02:42

HongChao6 force-pushed the getevent branch 3 times, most recently from 4b49bd9 to 9413590 Compare March 25, 2026 06:30

HongChao6 requested a review from cederom March 25, 2026 06:34

acassis reviewed Mar 25, 2026

View reviewed changes

graphics/input/Kconfig Outdated Show resolved Hide resolved

This comment was marked as outdated.

Sign in to view

acassis requested changes Mar 25, 2026

View reviewed changes

cederom requested changes Mar 25, 2026

View reviewed changes

xiaoxiang781216 previously approved these changes Mar 26, 2026

View reviewed changes

simbit18 previously approved these changes Mar 26, 2026

View reviewed changes

HongChao6 dismissed stale reviews from simbit18 and xiaoxiang781216 via 3796c6e March 27, 2026 06:58

HongChao6 force-pushed the getevent branch from 9413590 to 3796c6e Compare March 27, 2026 06:58

HongChao6 requested review from acassis, simbit18 and xiaoxiang781216 March 27, 2026 07:30

HongChao6 force-pushed the getevent branch from 3796c6e to 5bed08d Compare March 27, 2026 07:52

HongChao6 force-pushed the getevent branch 2 times, most recently from be70cf1 to d7ce4f6 Compare March 30, 2026 06:10

HongChao6 force-pushed the getevent branch from d7ce4f6 to 34df408 Compare March 31, 2026 11:48

linguini1 requested changes Mar 31, 2026

View reviewed changes

graphics/input/getevent/getevent.c Outdated Show resolved Hide resolved

graphics/input/getevent/getevent.c Outdated Show resolved Hide resolved

graphics/input/getevent/getevent.c Outdated Show resolved Hide resolved

HongChao6 force-pushed the getevent branch from 34df408 to 7692105 Compare April 1, 2026 02:38

HongChao6 requested a review from linguini1 April 1, 2026 08:03

HongChao6 force-pushed the getevent branch from 7692105 to b0b9bb5 Compare April 2, 2026 04:47

HongChao6 force-pushed the getevent branch from b0b9bb5 to ebc6f8f Compare April 2, 2026 11:37

simbit18 approved these changes Apr 2, 2026

View reviewed changes

linguini1 approved these changes Apr 2, 2026

View reviewed changes

Conversation

HongChao6 commented Mar 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

What

When / Why

Usage

Configuration

Impact

Testing

PR verification Self-Check

Uh oh!

cederom left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

HongChao6 commented Mar 25, 2026

Uh oh!

Uh oh!

This comment was marked as outdated.

Uh oh!

acassis left a comment

Choose a reason for hiding this comment

Uh oh!

cederom commented Mar 25, 2026

Uh oh!

acassis commented Mar 25, 2026

Uh oh!

cederom commented Mar 25, 2026

Uh oh!

cederom left a comment

Choose a reason for hiding this comment

Uh oh!

cederom commented Mar 25, 2026

Uh oh!

simbit18 commented Mar 27, 2026

Uh oh!

HongChao6 commented Mar 27, 2026

Uh oh!

simbit18 commented Mar 27, 2026

Uh oh!

HongChao6 commented Mar 30, 2026

Uh oh!

simbit18 commented Mar 30, 2026

Uh oh!

JianyuWang0623 commented Mar 30, 2026

Uh oh!

simbit18 commented Mar 30, 2026

Uh oh!

linguini1 commented Mar 30, 2026

Uh oh!

simbit18 commented Mar 31, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

linguini1 commented Mar 31, 2026

Uh oh!

HongChao6 commented Apr 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

linguini1 commented Apr 1, 2026

Uh oh!

linguini1 commented Apr 1, 2026

Uh oh!

HongChao6 commented Apr 2, 2026

Uh oh!

linguini1 commented Apr 2, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

HongChao6 commented Mar 25, 2026 •

edited

Loading

cederom left a comment •

edited

Loading

HongChao6 commented Apr 1, 2026 •

edited

Loading