Go to advanced search

by PhilE
Thu Feb 15, 2018 10:47 pm
Forum: Device Tree
Topic: i2c-gpio dtoverlay bus number
Replies: 4
Views: 137

Re: i2c-gpio dtoverlay bus number

I'm sure it will be possible - there are many board .dts files with multiple i2c-gpio instances - it is just a question of figuring out the bus numbering and whether we can use one standard overlay repeatedly or whether it must be a custom overlay.
by PhilE
Thu Feb 15, 2018 9:57 pm
Forum: Device Tree
Topic: i2c-gpio dtoverlay bus number
Replies: 4
Views: 137

Re: i2c-gpio dtoverlay bus number

Can you really not get the three I2C devices (or at least two of them) to share the hardware I2C bus? It would be more efficient.
by PhilE
Thu Feb 15, 2018 4:06 pm
Forum: Device Tree
Topic: Possibility for multiple instances of LIRC?
Replies: 9
Views: 203

Re: Possibility for multiple instances of LIRC?

One thing I just noticed is that the gpio-ir overlay is missing the (required) pinctrl-0 and (optional) pinctrl-names properties - the -tx overlays have them. Could you add these to gpio-ir-overlay.dts?
You're quite right - that's fixed now.
by PhilE
Thu Feb 15, 2018 2:09 pm
Forum: Device Tree
Topic: Possibility for multiple instances of LIRC?
Replies: 9
Views: 203

Re: Possibility for multiple instances of LIRC?

Thanks for covering the lircd daemon, HiassofT - it's not something I've played with yet other than in the most basic way. vicary - we like curiosity around these parts, so keep on asking questions. My reply assumed too much knowledge, so I'll try to correct that now. The change I've made should mak...
by PhilE
Thu Feb 15, 2018 9:53 am
Forum: Device Tree
Topic: Possibility for multiple instances of LIRC?
Replies: 9
Views: 203

Re: Possibility for multiple instances of LIRC?

HiassofT was right to steer you towards gpio-ir and gpio-ir-tx - lirc-rpi isn't built in the 4.15 kernel tree.

To make it easier to use multiple gpio-ir instances the overlays (gpio-ir and gpio-ir-tx) now use the magic "reg" property to prevent name clashes.
by PhilE
Mon Feb 12, 2018 4:43 pm
Forum: Device Tree
Topic: [Custom Kernel] after compilation dtb modules are missing
Replies: 19
Views: 3243

Re: [Custom Kernel] after compilation dtb modules are missing

What are you building on? I'm not near a computer, but you should at least include ARCH=arm in the make command line.
by PhilE
Thu Feb 08, 2018 9:57 am
Forum: Device Tree
Topic: external watchdog on GPIO output
Replies: 34
Views: 4567

Re: external watchdog on GPIO output

There are many reasons to want to use the existing watchdog interface: * It should be (marginally) less code. * Any code that uses it will be portable to other Linux systems because the API is separated from the hardware by multiple layers of abstraction - the fact that the watchdog is reached via G...
by PhilE
Thu Feb 08, 2018 9:51 am
Forum: Compute Module
Topic: CM with LAN chip - use RPi Foundation OUI
Replies: 2
Views: 262

Re: CM with LAN chip - use RPi Foundation OUI

Thanks for asking - we are happy for CM customers to use the Pi OUI with an Ethernet interface of their choice. The firmware writes the MAC address into the Device Tree if there is a node with the label "ethernet0", but if you are unable to make use of this mechanism then allocating MAC addresses by...
by PhilE
Wed Feb 07, 2018 11:31 am
Forum: Raspberry Pi Desktop for PC and Mac
Topic: Files Are Compiled As Shared Libraries Instead Of Executables
Replies: 18
Views: 707

Re: Files Are Compiled As Shared Libraries Instead Of Executables

My findings are slightly different, but I agree that Position Independent Executables confuse the mimetype mechanism. In the most recent Raspbian Stretch image (Nov 29th 2017) gcc generates normal position-dependent code by default, but I can believe that somewhere the default has changed. Adding '-...
by PhilE
Tue Feb 06, 2018 3:18 pm
Forum: Raspberry Pi Desktop for PC and Mac
Topic: Files Are Compiled As Shared Libraries Instead Of Executables
Replies: 18
Views: 707

Re: Files Are Compiled As Shared Libraries Instead Of Executables

Let's start from the basics: build a program with and without the -no-pie flag and post the output of "file <progname>" (where <progname> is the name of the executable) for both versions.
by PhilE
Fri Feb 02, 2018 5:27 pm
Forum: Interfacing (DSI, CSI, I2C, etc.)
Topic: One wire w1-gpio and w1-gpio-custom
Replies: 70
Views: 69753

Re: One wire w1-gpio and w1-gpio-custom

[ "modprobe w1-gpio" is unnecessary - the overlay will ensure the module is loaded. ] I'm sure your Pi would run out of some precious resource eventually if you were to load both overlays every minute. Why not check for the existence of the overlays before loading them, or load them in some other on...
by PhilE
Fri Feb 02, 2018 1:57 pm
Forum: Interfacing (DSI, CSI, I2C, etc.)
Topic: One wire w1-gpio and w1-gpio-custom
Replies: 70
Views: 69753

Re: One wire w1-gpio and w1-gpio-custom

They may be in /usr/bin (actually /opt/vc/bin - the others are symlinks) but they are text file scripts - just open them up in your favourite editor, provided you have root privileges or use sudo. After some internal testing we've merged a patch that makes this change, but also falls back to LOGNAME...
by PhilE
Fri Feb 02, 2018 9:27 am
Forum: Interfacing (DSI, CSI, I2C, etc.)
Topic: One wire w1-gpio and w1-gpio-custom
Replies: 70
Views: 69753

Re: One wire w1-gpio and w1-gpio-custom

These error messages are triggered by the dtoverlay-pre and dtoverlay-post scripts that allow certain actions to be taken before adding or removing overlays. These hooks were added to allow the lxpanel volume control to be told to close its ALSA connections, otherwise unloading soundcard overlays te...
by PhilE
Tue Jan 30, 2018 9:50 am
Forum: Troubleshooting
Topic: rpi-update fails when updating > 16 rpis concurrently
Replies: 4
Views: 201

Re: rpi-update fails when updating > 16 rpis concurrently

The correct way to avoid the GitHub rate limiting is to get your own GitHub API token - see here to create one, and here for how to use it with rpi-update. I'm assuming you've made an informed decision that rpi-update is the right tool for your upgrading needs - do not use for mission-critical appli...
by PhilE
Mon Jan 29, 2018 9:58 am
Forum: Compute Module
Topic: Setting GPIO alternate functions by modifying the device tree
Replies: 2
Views: 225

Re: Setting GPIO alternate functions by modifying the device tree

The dt-blob (it's not really Device Tree) file must be called "dt-blob.bin" and copied into the root of the SD card (/boot) alongside the other boot files. Also, pin 31 isn't TXD0 - it's RTS0. I think you want 32 and 33. There is now an easier way to achieve the same effect, and that is to use the n...
by PhilE
Fri Jan 26, 2018 4:31 pm
Forum: Device Tree
Topic: setting mac address through device tree
Replies: 1
Views: 237

Re: setting mac address through device tree

You've got two options: 1. Set the "mac-address" property in the same location as the "local-mac-address" property is found in the .dtb. The firmware overwrites "local-mac-address" but leaves "mac-address" (which is checked first) free in case the .dtb wants to override it. 2. Use the "force_mac_add...
by PhilE
Thu Jan 25, 2018 4:59 pm
Forum: Device Tree
Topic: using uart on RPI B+ with jessie
Replies: 6
Views: 383

Re: using uart on RPI B+ with jessie

A fresh install of the 2016-09-23 Jessie image (this matches your kernel) on a B+ gives me UART0 mapped to GPIOs 14 & 15. No configuration changes are needed. Adding "enable_uart=1" doesn't change anything.

If you have a spare card, try a clean install and verify that you get the same results.
by PhilE
Thu Jan 25, 2018 12:49 pm
Forum: Device Tree
Topic: using uart on RPI B+ with jessie
Replies: 6
Views: 383

Re: using uart on RPI B+ with jessie

The firmware and kernel packages shipped with Jessie should enable UART0 automatically on a B+ unless there is something in your config.txt preventing it. It might be worth uploading config.txt and the output from "dmesg" after booting.
by PhilE
Thu Jan 25, 2018 9:54 am
Forum: General discussion
Topic: Wireless firmware testers wanted
Replies: 45
Views: 3891

Re: Wireless firmware testers wanted

Code: Select all

sudo apt-get update
sudo apt-get install bluez-firmware
will get you the latest BT firmware, which may be a slight improvement over what you have.
by PhilE
Thu Jan 25, 2018 9:23 am
Forum: Device Tree
Topic: using uart on RPI B+ with jessie
Replies: 6
Views: 383

Re: using uart on RPI B+ with jessie

Which UART are you trying to use? There are two. * uart0 is a pl011, also known as /dev/ttyAMA0. * uart1 is an 8250 clone, also known as the mini-UART and /dev/ttyS0. UART0 is the natural UART to use on any Pi without on-board Bluetooth - with no config.txt settings it will be enabled by default (en...
by PhilE
Mon Jan 22, 2018 12:32 pm
Forum: Device Tree
Topic: mmc1 SDIO clock changing
Replies: 5
Views: 491

Re: mmc1 SDIO clock changing

A pair of patches have gone into the kernel tree to add a max-frequency support to both downstream MMC drivers and to support sd_overclock underclocking in bcm2835-sdhost.
by PhilE
Mon Jan 22, 2018 9:40 am
Forum: Device Tree
Topic: mmc1 SDIO clock changing
Replies: 5
Views: 491

Re: mmc1 SDIO clock changing

The overclock mechanism in the bcm2835-sdhost driver is more complicated because it will automatically back-off the overclock if it sees an error. To prevent this backing-off continuing indefinitely it gets cancelled when the overclock reaches 50MHz, i.e. no longer overclocked. Because of a concern ...
by PhilE
Sun Jan 21, 2018 3:48 pm
Forum: Device Tree
Topic: mmc1 SDIO clock changing
Replies: 5
Views: 491

Re: mmc1 SDIO clock changing

The DT bindings documentation for mmc says: ... - max-frequency: maximum operating clock frequency ... I believe you should be able to use this property to limit the frequency requested by the MMC subsystem. However, reading the driver suggests this value is ignored, being overwritten by the maximum...
by PhilE
Fri Jan 19, 2018 11:19 am
Forum: Compute Module
Topic: Rootfs in read-only: any advantage?
Replies: 4
Views: 373

Re: Rootfs in read-only: any advantage?

Marking the filing system as read-only is asking the OS not to write to the eMMC, while making the eMMC read-only would actually stop any writes from happening even if the OS tried.

Go to advanced search