Differences

This shows you the differences between two versions of the page.

--- doc:appunti:hardware:mini_pc_intel_pentium_j3710 [2026/02/22 18:02] – niccolo
+++ doc:appunti:hardware:mini_pc_intel_pentium_j3710 [2026/02/22 19:35] (current) – [Pygame] niccolo
@@ Line 1: / Line 1: @@
 ====== Mini PC Intel Pentium J3710 ======
-===== Video card problem =====
+===== Video card problem with kernel 6.12 =====
-Using **Debian 13 Trixie** with **kernel 6.12.69**, the screen goes blank (no HDMI signal) as soon as it switches to framebuffer. With dmesg you can read the following errors:
+Using **Debian 13 Trixie** with **kernel 6.12.69**, the screen goes blank (no HDMI signal) as soon as it switches to frame buffer. With **dmesg** you can read the following errors:
 <code>
@@ Line 19: / Line 19: @@
 The only way to boot Debian 13 with kernel 6.12 is to pass the **nomodeset** kernel parameter.
+===== Kernel 6.1 =====
+It is possible to install **kernel 6.1.0** from **Debian 12 Bookworm**, so that frame buffer works:
+<code>
+[    5.399065] fbcon: i915drmfb (fb0) is primary device
+[    5.438998] Console: switching to colour frame buffer device 128x37
+[    5.458533] i915 0000:00:02.0: [drm] fb0: i915drmfb frame buffer device
+</code>
+Despite the HDMI monitor is connected, the status pseudfile reports that it is **disconnected**:
+<code>
+cat /sys/class/drm/card0-HDMI-A-1/status
+disconnected
+</code>
+Writing the **on** string to the file changes the status to connected:
+<code>
+echo on > /sys/class/drm/card0-HDMI-A-1/status
+cat /sys/class/drm/card0-HDMI-A-1/status
+connected
+</code>
+but it causes a kernel trace:
+<code>
+[   69.315446] i915 0000:00:02.0: timed out waiting for [ENCODER:94:HDMI B] port ready: got 0xf, expected 0x0
+</code>
+When the **i915** kernel module is loaded, you can find five **DRM** (Direct Rendering Manager) devices managed by **card0**:
+<code>
+ls -1 /sys/class/drm/
+card0
+card0-DP-1
+card0-DP-2
+card0-HDMI-A-1
+card0-HDMI-A-2
+card0-HDMI-A-3
+renderD128
+version
+</code>
+It is possibile to do a **complete hardware reset** of the cards using the following command:
+<code>
+echo "0000:00:02.0" > /sys/bus/pci/drivers/i915/unbind
+sleep 1
+echo "0000:00:02.0" > /sys/bus/pci/drivers/i915/bind
+</code>
+After the //unbind// you can verify that the directory **/sys/class/drm/** is empty, it will be populated again after the //bind//.
+===== Blanking the Virtual Terminal =====
+When the host is started in frame buffer text console, the **setterm** command can be used to force the screen to go blank:
+<code>
+setterm --blank force
+setterm --powerdown 0
+</code>
+The monitor turns off due no signal and remains blank even keys are pressed on the keyboard.
+To revive the screen:
+<code>
+setterm --blank poke
+setterm --powerdown 0
+</code>
+If the screen is blanked the following command returns **1**:
+<code>
+setterm --blank
+</code>
+===== Pygame =====
+We verified what driver uses the Pygame app:
+<code python>
+import pygame
+pygame.init()
+print(pygame.display.get_driver())
+</code>
+It says **KMSDRM**.
+Using that driver the SDL **Simple DirectMedia Layer** library owns the DRM master, external tools (modetest, sysfs) do not work for blanking the screen. The correct solution is to control this from inside SDL.