So my children are using now one of my 10IVs. The 3,5mm-jack isn’t working (neither with youtube in browser nor with an audio in Fernschreiber nor while calling). Any idea how to investigate that?
Difficult to tell if it could be a hardware or software issue, but as nothing has changed in software I would initially suspect hardware.
I would say use a plastic or wooden toothpick gently to pick out any dust in the 3.5mm jack hole, and see if the CSD tool can detect the headphone or the headphone mic or buttons.
You could also try the audio output chooser app to override what the OS detects and send audio to the headphones no matter what.
I have updated. Only one fingerprint works. Stops working when i add second one. Everything as described. Thank you for adding fingerprint! How about cooperate with jolla to force Android App Support? I’am very struggling without it. I wish sailfish become more mass-marketed so to application availability isolation would be less felt! And i dream they to make native Rust SDK support in order to have QML also checked at compile time alongside of compiling of core of application too. The components of OS rewritten in Rust - would enhance stability dramatically too keeping performance same as C written apps do with having good zero-cost abstractions like Python does.
I am not a professional in configuring drivers. Only one thing that I could share is that it works.
Re: Fingerprint, glad it’s working. Not sure how to tackle adding a second Fingerprint yet, I don’t know why that doesn’t work.
Thanks for your work on Waydroid, I’m yet to try it on my phone and I’m afraid I’m not currently across the technical side of it enough to understand where you’ve hit a wall, but if I get the chance sometime in the next week or so I’m keen to give it a try. In the meantime I’m working on the 5.2 update.
Re: Jolla Appsupport, in the 03/09/26 meeting, we heard this from Jolla:
Since then I have reached out at least twice on the #sailfishos channel and not heard anything back. All I can do I guess is bring it up again in the next meeting, but I cannot attend meetings due to them occurring at 2AM my time.
Thank you for your involve. That’s gave me hope to use 10 IV as my daily driver.
Thanks for your work on Waydroid, I’m yet to try it on my phone and I’m afraid I’m not currently across the technical side of it enough to understand where you’ve hit a wall, but if I get the chance sometime in the next week or so I’m keen to give it a try. In the meantime I’m working on the 5.2 update.
What I would like to ask for
1. Kernel support for AAS / Waydroid. Please take a look at the kernel parts that are required to run AAS or Waydroid. During my AI agent investigation I hit a wall in some branch because of a missing kernel component, possibly something graphics-related, I do not remember exactly which.
2. My notes, if you want them. If you wish, I can send my updated AI notebook, where the agent wrote down its steps to get this at least working. Some of the lines there are totally trash, but with it the AI managed to make it work for the third time already, from scratch.
On the AI notebook
Most of the frustration in that work was not the missing component itself. It was the agent forgetting the main goal and deep diving into irrelevant details, which I assume might be avoided by higher-level thinking. That is a process problem, not a kernel problem, and I still think the notes are worth sharing.
Fingerprint
About the fingerprint: at least one of them is much more usable for me for now. I can live with that.
One more thing
It would be nice for the ecosystem if there was a way to offer the Jolla team funding to build a server cluster for AI models aimed at natural-language application production, so that developers and users could build applications by simply explaining the concept.
Right now I am doing something similar myself: an application in Rust + QML + an opencode server that runs completely on my device, with voice input for requests and voice synthesis for the AI, so I can plan and build an application by conversation, in my free time.
AFAIK I have enabled the kernel configs required, but I’m not an expert on Waydroid / AAS. If you can work out which specifics are missing I am more than willing to enable them and rebuild the kernel ![]()
To Zambezi and Murray users: 5.2 update installed tonight on my phone, seems to run very well. Solved some video decoding issues in the process.
I shall release the 5.2 update ASAP, pending Jolla updating the repositories for it.
I am not going to be promising. But if I have free time, AI tokens and the will Then I will investigate.
Thanks! I should have done your suggestions on my own before posting - say three times out loud “I’m not a newbie, I’m not a newbie, I’m not a newbie!”
So, CSD tool fails about the headphone. But audio output chooser works and I have music on headphone! So, perhaps I killed something, when I tried to install AAS on the phone. I think, except of that I did nothing special with the phone - or could it be, that there is a difference with XQ-CC54?
Funny thing: Even while I’m listening to music over the headphone, CSD-tools are still complaining about “Headset not detected”.
Now I’m playing around with the phone more - your port is incredibly fast! Its such a joy to use this phone.
It sounds to me like there is dust in the bottom of the headphone jack, this will stop the device detecting the headphone has been connected. It is a common issue with headphone jacks and has come up a few times on this forum before. At any rate, you can continue to use the audio output chooser app if you wish, but I would advise trying to gently remove the dust.
Thankyou, I’m very glad you’re enjoying it.
But—it’s been working for me since today! Xperia 10IV XQQC-54… Thanks to @logidroid…
![]()
That file does not exist on this port, and headphone jack has worked from the start, I haven’t changed anything.
In 5.2 there are a few changes to audio stuff but nobody has that yet, only me and juz. The latest that anybody has access to still has all the same audio configuration as when the port was first released.
Oh right, it also depends on the version of the Sony drivers the person is using, because in the version where the headphone jack works, mobile networking is completely broken.
Nah, its not the dust. I cleaned it. But my children are on the way with this phone, so no further investigation possible in the next days.
But second XQCC54 is flashed now (it had the official port running till tonight) and the jack is working.
Perhaps interesting: At first try I was curious about the Lineage-Version and flashed the latest LineageOS 23.2 and over that your SFOS-Port. It didn’t boot. So its important to use LineageOS 22.
First thing i have remembered was missing to start mine community port to adopt Android App support - was missing flag in kernel that includes support of squash file system to be mounted. I was doing the trick that made my community port while connecting to Jolla store of applications it to present itself not like XQ-CC72 but like XQ-CC54 - for which i have the native license for AAS. Then after launching store - Android App Support - did appeared in the store to download. But installation was failing for some reason. When i asked AI to install parts from image of official port ant the rest from the repo - then i was getting that error of missing Squash FS support by kernel.
Here is what i have got from other developer who makes something similar to Android App Support (I’am not sure i I’am allowed to reveal all the details about the software that man makes) but here what he sent me to share to be supported:
| option | purpose |
|---|---|
CONFIG_NAMESPACES, CONFIG_PID_NS, CONFIG_NET_NS, CONFIG_IPC_NS |
container namespaces |
CONFIG_CGROUPS |
control groups |
CONFIG_CGROUP_DEVICE or CONFIG_CGROUP_BPF |
device access control — one is enough (cgroup v1 or v2; recent SailfishOS uses v2) |
CONFIG_OVERLAY_FS |
overlaying the Android images |
CONFIG_ANDROID_BINDER_IPC, CONFIG_ANDROID_BINDERFS |
binder, with nodes created at runtime |
CONFIG_BRIDGE, CONFIG_VETH, CONFIG_TUN |
container networking |
CONFIG_NETFILTER_ADVANCED, CONFIG_NETFILTER_XTABLES |
iptables |
CONFIG_NETFILTER_XT_TARGET_CHECKSUM |
without it, the container’s DHCP gets no replies |
CONFIG_IP_NF_MANGLE, CONFIG_IP_NF_NAT |
NAT towards the host |
Kernel — recommended, not blocking
CONFIG_SQUASHFS,CONFIG_CGROUP_PIDS,CONFIG_IP6_NF_MANGLE.CONFIG_ASHMEM— missing on recent kernels, which is normal (replaced by memfd).
$ zcat /proc/config.gz | grep -iEw "CONFIG_NAMESPACES|CONFIG_PID_NS|CONFIG_NET_NS|CONFIG_IPC_NS|CONFIG_CGROUPS|CONFIG_CGROUP_DEVICE|CONFIG_CGROUP_BPF|CONFIG_OVERLAY_FS|CONFIG_ANDROID_BINDER_IPC|CONFIG_ANDROID_BINDERFS |CONFIG_BRIDGE|CONFIG_VETH|CONFIG_TUN|CONFIG_NETFILTER_ADVANCED|CONFIG_NETFILTER_XTABLES|CONFIG_NETFILTER_XT_TARGET_CHECKSUM|CONFIG_IP_NF_MANGLE|CONFIG_IP_NF_NAT|CONFIG_SQUASHFS|CONFIG_CGROUP_PIDS|CONFIG_IP6_NF_MANGLE|CONFIG_ASHMEM"
CONFIG_CGROUPS=y
# CONFIG_CGROUP_PIDS is not set
CONFIG_CGROUP_DEVICE=y
CONFIG_CGROUP_BPF=y
CONFIG_NAMESPACES=y
CONFIG_IPC_NS=y
CONFIG_PID_NS=y
CONFIG_NET_NS=y
CONFIG_NETFILTER_ADVANCED=y
CONFIG_NETFILTER_XTABLES=y
CONFIG_NETFILTER_XT_TARGET_CHECKSUM=y
CONFIG_IP_NF_NAT=y
CONFIG_IP_NF_MANGLE=y
CONFIG_IP6_NF_MANGLE=y
CONFIG_BRIDGE=y
CONFIG_TUN=y
CONFIG_VETH=y
CONFIG_ASHMEM=y
CONFIG_ANDROID_BINDER_IPC=y
CONFIG_ANDROID_BINDERFS=y
CONFIG_OVERLAY_FS=y
CONFIG_SQUASHFS=y
The only one Murray and Zambezi is missing is CONFIG_CGROUP_PIDS, which you said is not blocking?
SquashFS is certainly enabled.
Update on Murray and Zambezi
Over the last month there has been some exciting changes to these ports.
First and foremost, I have been invited to join the prestigious Verdanditeam and bring these ports along with me. Working with @Mister_Magister, these ports are progressing in leaps and bounds.
I have moved my efforts to the 5.2 release. 5.2 is running very well on Murray and Zambezi, and will be in user’s hands as soon as Jolla releases it to devices other than the JP26. As part of the 5.2 release, many fixes are coming to Murray and Zambezi:
- Improved video codec support, fixing playback issues on the Youtube website and with some local videos
- Rejecting a phone call without answering it is now solved
- All four cameras are fully functional, both for photography and video recording.
Unfortunately, @juz and I have experienced a regression in battery life while using GPS under 5.2, this is a known issue that I have added to the list of things to work on.
There are also Verdanditeam QoL improvements to come, including:
- GUI updating from the Settings app
- Optional pre-installed 3rd party apps on fresh installs to make getting started easier (Chum, Storeman, PUNKY suite, etc.)
Overall, I’m super thrilled with the progress we’re still making, and very keen to be a part of Verdanditeam. Keep an eye on this thread, further improvements will surely follow!




