Librem5 issueshttps://source.puri.sm/groups/Librem5/-/issues2020-03-20T11:01:26Zhttps://source.puri.sm/Librem5/OS-issues/-/issues/143PUK dialog failure results in PIN dialog2020-03-20T11:01:26ZDorota CzaplejewiczPUK dialog failure results in PIN dialog# What problem did you encounter
After entering an incorrect PUK, the PIN dialog is shown.
## What is the actual behaviour?
1. Fail unlocking the PIN until PUK dialog appears with X attempts left
2. Enter an invalid 8-digit number: PI...# What problem did you encounter
After entering an incorrect PUK, the PIN dialog is shown.
## What is the actual behaviour?
1. Fail unlocking the PIN until PUK dialog appears with X attempts left
2. Enter an invalid 8-digit number: PIN is requested
## What is the expected behaviour?
2. Enter an invalid 8-digit number: PUK dialog appears with a reduced attempt count
# What hardware are you running on?
L5 Birch
# Relevant OS information
## Any other information that may be helpful?
OS last updated about 2020-03-01https://source.puri.sm/Librem5/debs/ModemManager/-/issues/34Submitting udev rule upstream?2020-06-29T20:07:34ZClayton CraftSubmitting udev rule upstream?I am using upstream ModemManager (the latest release, which includes the series submitted by @bob.ham) on postmarketOS with Phosh/Calls, and placing/receiving phone calls was really unreliable for me. I needed to include the udev rule in...I am using upstream ModemManager (the latest release, which includes the series submitted by @bob.ham) on postmarketOS with Phosh/Calls, and placing/receiving phone calls was really unreliable for me. I needed to include the udev rule in this patch for things to start working: 27113daef1af7c2767d0a48d2ae6c5176d1c8f3e
Is there any reason why this wasn't submitted upstream?https://source.puri.sm/Librem5/linux/-/issues/178devfreq: use depending on bandwidth requirements2021-01-11T16:46:28ZMartin Kepplingerdevfreq: use depending on bandwidth requirementsThe NXP kernels' busfreq subsystem switches frequencies depending on requirements that various other subsystem require / release from it. We should do the same, in order to eventually get rid of our `power_key.py` userspace settings.
ht...The NXP kernels' busfreq subsystem switches frequencies depending on requirements that various other subsystem require / release from it. We should do the same, in order to eventually get rid of our `power_key.py` userspace settings.
https://source.puri.sm/Librem5/use-cases/issues/75https://source.puri.sm/Librem5/OS-issues/-/issues/144Connection activation fails when SIM is locked on boot2020-03-31T14:01:27ZArnaud FerrarisConnection activation fails when SIM is locked on bootWhen booting with a PIN-locked SIM card, the mobile data connection is not activated upon unlocking the SIM: the moment the modem is registered in ModemManager, two prompts show up:
1. a functional system-prompt triggered by gnome-settin...When booting with a PIN-locked SIM card, the mobile data connection is not activated upon unlocking the SIM: the moment the modem is registered in ModemManager, two prompts show up:
1. a functional system-prompt triggered by gnome-settings-daemon, which can be used to enter the PIN code and unlock the SIM
2. an [empty network-auth-prompt](https://gitlab.com/a-wai/debos-pinephone/-/issues/5) triggered by NetworkManager on which only the Cancel button can be selected
This happens on each boot once a mobile data connection has been setup.
Based on my tests, here's what happens:
1. The modem is registered by ModemManager, and the corresponding network interface (wwan0) is created, even though the SIM is locked
2. gsd is notified that a modem is present, and that the SIM is locked -> first prompt is displayed
2a. NetworkManager sees the wwan0 device, tries to activate the data connection, but can't because the SIM is locked -> it sends a "secrets request" for the PIN code, displaying the second prompt (which is empty because only Wifi networks are supported for now)
3. As the network-auth-prompt is not functional, the user cancels it; NetworkManager subsequently fails to activate the data connection
4. The PIN is entered in the system-prompt, SIM is then unlocked; however, as the connection activation already failed once, NetworkManager doesn't attempt to activate it again
5. The connection can be activated manually using `nmcli` or disabling then enabling it in Settings
This raises the question of handling locked SIM cards, as:
- on first boot, the mobile data connection doesn't exist, so gsd unlocking the SIM looks like the only way to go
- however, on subsequent boots, if data connection has been setup previously, there's this sort of "race condition" where NM shouldn't try to bring up a connection on a locked modem but still does
FYI, I have prototyped having the network-auth-prompt [work for PIN requests](https://gitlab.com/a-wai/phosh/-/commit/370dff263010904f4a4b04e9491bfcbc6cef1cd8), but I'm filing this issue to start a discussion about this whole unlocking/connection process, as I believe it can be more complicated to sort out efficiently than it seems.
("NM should wait for the SIM to be unlocked" could be a valid answer, but I'm not aware of NM internals and how it interacts with ModemManager exactly)
Note: This has been verified on a PinePhone running a Debian-based image (custom built packages, phosh v0.2.1) and an unofficial PureOS build (untouched PureOS packages)https://source.puri.sm/Librem5/linux/-/issues/179Serial Over USB No Longer Working (no Linux USB Gadget shows up)2021-09-07T16:31:48ZEric KuzmenkoSerial Over USB No Longer Working (no Linux USB Gadget shows up)I tested 5.6.0-rc7 8adc003c and nothing shows up, in SDP mode it always shows up.
After trying 5.6.0-rc7 I reverted back to 5.3.0 built by Jenkins and it works every time, the one built here: https://arm01.puri.sm/job/Kernel_builds/job/...I tested 5.6.0-rc7 8adc003c and nothing shows up, in SDP mode it always shows up.
After trying 5.6.0-rc7 I reverted back to 5.3.0 built by Jenkins and it works every time, the one built here: https://arm01.puri.sm/job/Kernel_builds/job/kernel_linux-next_imx8-current-librem5/9/
A previous kernel I tested that was built on March 13th was working sometimes but it seems to have gotten even worse where it doesn't work at all now. So I believe something within the month of March broke this.https://source.puri.sm/Librem5/developer.puri.sm/-/issues/155Select a translation tool/service/process for the documentation2020-09-15T16:48:41ZDavid BoddieSelect a translation tool/service/process for the documentationThis should be aligned with anything used for Librem One and other documentation. See #6 for a discussion of the tools and mechanisms used to integrate Sphinx with whatever gets chosen.This should be aligned with anything used for Librem One and other documentation. See #6 for a discussion of the tools and mechanisms used to integrate Sphinx with whatever gets chosen.https://source.puri.sm/Librem5/quickstart-guide/-/issues/20First start confusions2020-03-26T15:21:25ZDorota CzaplejewiczFirst start confusionsI'm going through the quick start, and here's what I'm noticing that I'm finding a little confusing.
I will take into account that the quick start online reflects the state of the software that will be shipped next rather than what has ...I'm going through the quick start, and here's what I'm noticing that I'm finding a little confusing.
I will take into account that the quick start online reflects the state of the software that will be shipped next rather than what has been shipped already, although this is a problem when the paper version is confusing and the user wants to reach the online one for clarifications. Perhaps the paper version should refer to some "release branch" online that we can clarify while still describing what was actually shipped?
As for general observations (I won't mention things fixed in the online version over paper):
- the guide tells me that a SIM card is required (debatable)
- the need for first charge is indicated (great)
- the guide mentions a charge indicator, which is not there (confusing)
- the guide doesn't explain what should happen to the phone when it's plugged in and successfully charging: in my case the LED goes into a complicated periodic pattern, but I have no idea if it's intended
- the guide explains how long the power button should be held (good)https://source.puri.sm/Librem5/linux/-/issues/180next: noc not probed correctly2020-08-29T10:43:15ZGuido Gunthernext: noc not probed correctlyI've just seen
```
# cat /sys/kernel/debug/devices_deferred
32700000.noc
```
so the driver asked for probe deferral. This does not always seem to happen but when it happens it breaks clocking down DRAM.I've just seen
```
# cat /sys/kernel/debug/devices_deferred
32700000.noc
```
so the driver asked for probe deferral. This does not always seem to happen but when it happens it breaks clocking down DRAM.https://source.puri.sm/Librem5/developer.puri.sm/-/issues/156Move How To guides from the Development Environment section2020-03-27T13:54:32ZDavid BoddieMove How To guides from the Development Environment sectionThe How To and Troubleshooting guides cover some low level features of the Librem 5 hardware for the devkit and phone. Some, but not all, of these should be moved to a more suitable place, perhaps in the Software Reference section.The How To and Troubleshooting guides cover some low level features of the Librem 5 hardware for the devkit and phone. Some, but not all, of these should be moved to a more suitable place, perhaps in the Software Reference section.https://source.puri.sm/Librem5/OS-issues/-/issues/145MMS breaks receiving messages from the sender.2021-06-26T12:54:08ZDavid HamnerMMS breaks receiving messages from the sender.It seems if someone sends you an image via text, you can no longer receive messages from that person. But you can still send them messages.
Step to reproduce:
* Verify you can text someone
* Send an image via MMS to the Librem 5
No me...It seems if someone sends you an image via text, you can no longer receive messages from that person. But you can still send them messages.
Step to reproduce:
* Verify you can text someone
* Send an image via MMS to the Librem 5
No message will be received for the MMS, and no subsequent messages will be received.Evghttps://source.puri.sm/Librem5/debs/ModemManager/-/issues/35Get rid of periodic signal quality checks2020-12-05T18:09:44ZSebastian KrzyszkowiakGet rid of periodic signal quality checksCurrently, ModemManager seems to poll the signal strength and access technology unconditionally every 30 seconds. This isn't great for battery life, as each check wakes up the modem from sleep and is generally wasted if the device just s...Currently, ModemManager seems to poll the signal strength and access technology unconditionally every 30 seconds. This isn't great for battery life, as each check wakes up the modem from sleep and is generally wasted if the device just sits there with screen off.
Investigate whether it's possible to make ModemManager listen to unsolicited messages from modem instead of resorting to polling.https://source.puri.sm/Librem5/linux/-/issues/181no GPS fix using gpsd2020-07-21T07:12:31ZMartin Kepplingerno GPS fix using gpsdwhen installing `gpsd-clients`, setting `/dev/gnss0` in `/etc/default/gpsd` and starting `cgps`, I don't get a fix when being outside.when installing `gpsd-clients`, setting `/dev/gnss0` in `/etc/default/gpsd` and starting `cgps`, I don't get a fix when being outside.https://source.puri.sm/Librem5/debs/ModemManager/-/issues/36Broadmobie BM818 Does not expose USSD interface2020-07-14T15:14:10ZGuido GuntherBroadmobie BM818 Does not expose USSD interfaceI wanted to use ppm to check my prepaid balance but it did not find MMs USSD interface:
on the l5:
```
gdbus introspect --system --dest org.freedesktop.ModemManager1 --object-path /org/freedesktop/ModemManager1/Modem/0 | grep -i ussd
...I wanted to use ppm to check my prepaid balance but it did not find MMs USSD interface:
on the l5:
```
gdbus introspect --system --dest org.freedesktop.ModemManager1 --object-path /org/freedesktop/ModemManager1/Modem/0 | grep -i ussd
[nothing]
```
but on my laptop
```
gdbus introspect --system --dest org.freedesktop.ModemManager1 --object-path /org/freedesktop/ModemManager1/Modem/0 | grep -i -A12 ussd
interface org.freedesktop.ModemManager1.Modem.Modem3gpp.Ussd {
methods:
Initiate(in s command,
out s reply);
Respond(in s response,
out s reply);
Cancel();
signals:
properties:
readonly u State = 1;
readonly s NetworkNotification = '';
readonly s NetworkRequest = '';
};
```https://source.puri.sm/Librem5/linux/-/issues/182replace legacy g_multi usb gadget module with configfs setup2021-09-01T09:04:15ZMartin Kepplingerreplace legacy g_multi usb gadget module with configfs setupIf only to get rid of unneeded, legacy kernel code, we could setup the gadget devices we need via configfs (see https://gitlab.com/postmarketOS/pmaports/-/blob/master/main/postmarketos-mkinitfs/init_functions.sh#L281 for show postmarketo...If only to get rid of unneeded, legacy kernel code, we could setup the gadget devices we need via configfs (see https://gitlab.com/postmarketOS/pmaports/-/blob/master/main/postmarketos-mkinitfs/init_functions.sh#L281 for show postmarketos does it, as an example) instead of including g_multi (and other legacy gadget modules).https://source.puri.sm/Librem5/redpine-firmware-nonfree/-/issues/1build_redpine_modules.sh fails on 5.6.0-1-librem52020-08-27T07:21:57ZDavid Heidelbergbuild_redpine_modules.sh fails on 5.6.0-1-librem5```
root@pureos:/tmp/redpine-firmware-nonfree# ./build_redpine_modules.sh
+ RSI_VER=RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5
++ pwd
+ CWD=/tmp/redpine-firmware-nonfree
++ uname -r
+ KERNEL_REL=5.6.0-1-librem5
+ HEADER_VERSION=linux-hea...```
root@pureos:/tmp/redpine-firmware-nonfree# ./build_redpine_modules.sh
+ RSI_VER=RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5
++ pwd
+ CWD=/tmp/redpine-firmware-nonfree
++ uname -r
+ KERNEL_REL=5.6.0-1-librem5
+ HEADER_VERSION=linux-headers-5.6.0-1-librem5
+ OUTPUT_DIR=/tmp/redpine-firmware-nonfree/linux-5.6.0-1-librem5
+ ARCH=arm64
+ '[' '!' -e /usr/src/linux-headers-5.6.0-1-librem5 ']'
+ '[' '!' -e /tmp/redpine-firmware-nonfree/linux-5.6.0-1-librem5 ']'
+ mkdir -p /tmp/redpine-firmware-nonfree/linux-5.6.0-1-librem5
+ tar -xf files/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5.tgz
+ RSI_BUILD_PATH=RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi/
++ echo /usr/src/linux-headers-5.6.0-1-librem5
++ sed 's/\//\\\//g'
+ ESCAPED_KERNEL_DIR='\/usr\/src\/linux-headers-5.6.0-1-librem5'
+ sed 's/KERNELDIR=.*/KERNELDIR=\/usr\/src\/linux-headers-5.6.0-1-librem5/' -i RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi//Makefile
+ sed 's/-Wall -Werror//' -i RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi//Makefile
+ cp files/update_redpine.sh /tmp/redpine-firmware-nonfree/linux-5.6.0-1-librem5
+ cp files/update_redpine.sh /tmp/redpine-firmware-nonfree/files/firmware
+ cd RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi/
+ echo '#define UTS_RELEASE "5.6.0-1-librem5"'
+ sudo tee /lib/modules/5.6.0-1-librem5/build/include/generated/utsrelease.h
#define UTS_RELEASE "5.6.0-1-librem5"
+ make ARCH=arm64 -C /lib/modules/5.6.0-1-librem5/build M=/tmp/redpine-firmware-nonfree/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi modules
make: Entering directory '/usr/src/linux-headers-5.6.0-1-librem5'
CC [M] /tmp/redpine-firmware-nonfree/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi/rsi_91x_sdio_ops.o
CC [M] /tmp/redpine-firmware-nonfree/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi/rsi_91x_sdio.o
LD [M] /tmp/redpine-firmware-nonfree/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi/rsi_sdio.o
CC [M] /tmp/redpine-firmware-nonfree/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi/rsi_91x_usb_ops.o
CC [M] /tmp/redpine-firmware-nonfree/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi/rsi_91x_usb.o
LD [M] /tmp/redpine-firmware-nonfree/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi/rsi_usb.o
CC [M] /tmp/redpine-firmware-nonfree/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi/rsi_91x_hal.o
CC [M] /tmp/redpine-firmware-nonfree/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi/rsi_91x_main.o
CC [M] /tmp/redpine-firmware-nonfree/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi/rsi_91x_mac80211.o
/tmp/redpine-firmware-nonfree/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi/rsi_91x_mac80211.c:2964:30: error: initialization of ‘int (*)(struct ieee80211_hw *, struct ieee80211_vif *)’ from incompatible pointer type ‘int (*)(struct ieee80211_hw *)’ [-Werror=incompatible-pointer-types]
.cancel_remain_on_channel = rsi_mac80211_cancel_roc,
^~~~~~~~~~~~~~~~~~~~~~~
/tmp/redpine-firmware-nonfree/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi/rsi_91x_mac80211.c:2964:30: note: (near initialization for ‘mac80211_ops.cancel_remain_on_channel’)
cc1: some warnings being treated as errors
make[1]: *** [scripts/Makefile.build:268: /tmp/redpine-firmware-nonfree/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi/rsi_91x_mac80211.o] Error 1
make: *** [Makefile:1683: /tmp/redpine-firmware-nonfree/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi] Error 2
make: Leaving directory '/usr/src/linux-headers-5.6.0-1-librem5'
+ cd /tmp/redpine-firmware-nonfree/RS9116.NB0.NL.LNX.PURISM_FW_UPGRADE.1.0.5/rsi/
+ tar -zcf /tmp/redpine-firmware-nonfree/linux-5.6.0-1-librem5/rsi_upgrade.tgz './*.ko'
tar: ./*.ko: Cannot stat: No such file or directory
tar: Exiting with failure status due to previous errors
+ cd /tmp/redpine-firmware-nonfree/files/firmware
+ tar -zcf /tmp/redpine-firmware-nonfree/linux-5.6.0-1-librem5/rsi_firmware.tgz ./RS9116_NLINK_WLAN_BT_IMAGE.FW1_0_5.rps ./RS9116_NLINK_WLAN_BT_IMAGE.FW1_2_1.rps ./RS9116_NLINK_WLAN_BT_IMAGE.FW1_2_20.rps ./RS9116_NLINK_WLAN_BT_IMAGE.rps ./update_redpine.sh
+ cd /tmp/redpine-firmware-nonfree
```https://source.puri.sm/Librem5/OS-issues/-/issues/146Shutdown when long-pressing the Power Button2020-06-22T12:17:35ZMartin KepplingerShutdown when long-pressing the Power ButtonWe actually already do this, but out-of-band in librem5-devkit-tools' power_key.py script. In order to get rid of that script in the long run, could this be done in phoc?We actually already do this, but out-of-band in librem5-devkit-tools' power_key.py script. In order to get rid of that script in the long run, could this be done in phoc?https://source.puri.sm/Librem5/linux/-/issues/183Etnaviv MMU faults, broken state after recovering hung GPU2021-08-19T00:46:46ZSebastian KrzyszkowiakEtnaviv MMU faults, broken state after recovering hung GPUSometimes when playing with OpenGL-heavy applications, I get hung GPU due to MMU faults. The driver thinks that the GPU has been recovered, but it's not true - the screen stays frozen until a reboot. When that happens, the system general...Sometimes when playing with OpenGL-heavy applications, I get hung GPU due to MMU faults. The driver thinks that the GPU has been recovered, but it's not true - the screen stays frozen until a reboot. When that happens, the system generally works, but some parts of it are unusable - for instance, phoc doesn't blank the screen anymore and trying to restart it just hangs.
```
[ 8891.411193] etnaviv-gpu 38000000.gpu: MMU fault status 0x00000002
[ 8891.417307] etnaviv-gpu 38000000.gpu: MMU 0 fault addr 0xfe472000
[ 8892.567778] device: 'devcd1': device_add
[ 8892.567851] PM: Adding info for No Bus:devcd1
[ 8892.567952] etnaviv-gpu 38000000.gpu: recover hung GPU!
```
Seems like the easiest way to trigger it is to run Pure Maps and just play with it for a few minutes.
Right now I'm seeing this on the 5.6 kernel; I don't remember whether it happened on 5.3 as well or not.
Initially mentioned in https://source.puri.sm/Librem5/linux-next/issues/151#note_94840https://source.puri.sm/Librem5/purple-mm-sms/-/issues/11Pidgin crash after send or receive a sms2020-04-06T05:50:02ZViktor Ivan Rodriguez AbdalaPidgin crash after send or receive a smsHi, when I receive a new sms the pidgin application close suddenly, I start again pidgin and ask for root password for each sms received, if a have 15 sms I need to write 15 times the root password, and when I write a new sms then pidgin...Hi, when I receive a new sms the pidgin application close suddenly, I start again pidgin and ask for root password for each sms received, if a have 15 sms I need to write 15 times the root password, and when I write a new sms then pidgin close again.https://source.puri.sm/Librem5/dvk-mx8m-bsb/-/issues/8Cellular modem hardware kill switch activation2020-04-06T18:25:57ZTodd WeaverCellular modem hardware kill switch activationReproduction:
1. Cellular ON
2. Cellular turned OFF
3. Active connection still visible
![hks-cellular-not-triggered](/uploads/67496ca3ac788673a9c23813c9d48b84/hks-cellular-not-triggered.webm)Reproduction:
1. Cellular ON
2. Cellular turned OFF
3. Active connection still visible
![hks-cellular-not-triggered](/uploads/67496ca3ac788673a9c23813c9d48b84/hks-cellular-not-triggered.webm)https://source.puri.sm/Librem5/linux/-/issues/184irq/43-mmc1 firing despite of engaged WiFi killswitch2023-06-02T19:51:25ZSebastian Krzyszkowiakirq/43-mmc1 firing despite of engaged WiFi killswitchWith WiFi and modem both off, powertop reports `irq/43-mmc1` (hwirq 23) as one of the most frequent wake-up reasons on an idle system:
```
Usage Events/s Category Description
3.2 ms/s 31...With WiFi and modem both off, powertop reports `irq/43-mmc1` (hwirq 23) as one of the most frequent wake-up reasons on an idle system:
```
Usage Events/s Category Description
3.2 ms/s 31.8 Process [irq/43-mmc1]
```
AFAIK mmc1 is the WiFi's SDIO interface.
The first suspicion was on `broken-cd` clause in device tree, but applying !112 does not help.