DaVinci Resolve Free: insufficient-space error on Aurora composefs root despite 531 GB free in /var

[GodOrigem]

I cannot install DaVinci Resolve Free through LinuxToys 7.7.6 on Aurora 44 (Kinoite), image 44.20261022.1, variant aurora-dx.

I installed LinuxToys using the official installer (curl -fsSL https://linux.toys/install.sh | bash), which downloaded the AppImage and integrated it with Gear Lever.

After selecting Resolve Free, installation stops with:

Espaço de armazenamento disponível insuficiente para a operação.\n/ 'root' -- 12 GB

This means “Insufficient storage space available for the operation.” The literal \n appears in the terminal.

My storage layout is below, with the LUKS identifier redacted:

Filesystem                   Size  Used Avail Use% Mounted on
composefs                     47M   47M     0 100% /
/dev/mapper/luks-<redacted>   929G  397G  531G  43% /var
tmpfs                         20G 1021M   19G   6% /tmp
/dev/mapper/luks-<redacted>   929G  397G  531G  43% /var/home

My / reports zero available space, but the writable volume has 531 GB free. /var and /var/home are on the same volume. My storage uses LUKS encryption; I have not tested an unencrypted installation or another machine, so I do not know whether encryption affects reproduction.

The inspected source code contains a matching explanation: check_disk_space() requires 12 GB free in both $HOME and /. The Atomic path calls this check before setting up Davincibox. This appears to reject the composefs root before container setup, regardless of the available writable storage. The source inspection refers to the linked commit, not the contents of my installed AppImage.

I expected the Atomic installation path to check the actual download, extraction, and container storage destinations and distinguish insufficient space from a read-only filesystem. The attempt stops before the script queries or downloads the latest Resolve version.

System information

Component	Details
Distribution	Aurora 44, based on Fedora Kinoite
Image version	44.20261022.1
Variant / image ID	aurora-dx
Release channel	stable
Version codename	Stargazer
OSTree version	44.20261022.1
Image build ID	a0913d7
Kernel	7.2.7-200.fc44.x86_64 (64-bit)
KDE Plasma	6.7.5
KDE Frameworks	6.31.0
Qt	6.11.2
Graphics session	Wayland
CPU	AMD Ryzen 5 5600X, 6 cores / 12 threads
RAM	40 GiB installed, 39.1 GiB usable
GPU	Intel Arc B580, dedicated; PCI ID shown by lspci: 07:00.0 Intel Corporation Battlemage G21 [Arc B580]
Motherboard	Gigabyte B550M AORUS ELITE
LinuxToys	7.7.6, AppImage installed by the official quick installer and integrated with Gear Lever
Resolve edition	Free
Resolve version	Not resolved: the failure occurs before the inspected script queries the latest stable version
Message language	Portuguese
Date observed	October 26, 2026

Relevant characteristics of this setup

• This is an image-based Atomic/OSTree system, with / provided by composefs, rather than a conventional writable root filesystem.
• / reports 47 MB total and zero available, while the writable storage mounted at /var and /var/home reports 531 GB available.
• The writable volume uses LUKS encryption. I have not compared an unencrypted setup, so its effect on reproduction is unknown.
• I am using the aurora-dx variant and the AppImage distribution of LinuxToys, rather than a layered RPM installation.
• The graphics environment is Wayland with a dedicated Intel Arc B580. No Resolve launch, OpenCL, or GPU processing error was observed; these details are context, not established causes.

The source check against / matches the failure observed on this setup. I have not verified reproduction on another Aurora image, another machine, or a different storage layout. Exact GPU driver versions were not collected.

I am posting here because direct issue creation is restricted. An attempted submission through the official reporting client failed at bug.linux.toys/bootstrap with HTTP 530; a separate service check returned Cloudflare error 1033. No issue was created by that attempt.

[psygreg]

There was an issue with the bug API due to a power loss on the server that has been already rectified. A fix will be provided soon.