gpo.zugaina.org

Search Portage & Overlays:

app-misc/greenboost

Tiered VRAM/RAM/NVMe memory pool for CUDA workloads

Screenshots

  • greenboost-3.4
    ~amd64
    abi_x86_32 bpf dist-kernel modules-compress modules-sign +strip

    View      Download      Browse     License: MIT   
    Overlay: hossie
  • greenboost-3.4
    ~amd64
    dist-kernel modules-compress modules-sign +strip

    View      Download      Browse     License: GPL-2   
    Overlay: tatsh-overlay

ChangeLog

commit 209c275b4821ec6eedcd24ed3077dfe087791e12
Author: Andrew Udvare <audvare@gmail.com>
Date: Mon Oct 12 14:36:30 2026 -0400

*/*: drop trailing period from DESCRIPTION

Signed-off-by: Andrew Udvare <audvare@gmail.com>

commit bd2f6e5aa60240f5085fbde076f12d605a6d92cb
Author: Andrew Udvare <audvare@gmail.com>
Date: Tue Sep 1 11:47:26 2026 -0400

app-misc/greenboost: add 3.4, drop 3.3

Signed-off-by: Andrew Udvare <audvare@gmail.com>

commit 83d2fbac55aac68a5d0f7a52be1e34d1dd5de8b4
Author: Andrew Udvare <audvare@gmail.com>
Date: Mon Aug 3 10:22:08 2026 -0400

app-misc/greenboost: rename from nvidia-greenboost, add 3.3

Upstream calls the project greenboost, so MY_PN goes away and the module,
dracut snippet and everything else follow $.

sci-ml/ollama has no cuda flag, so the USE dependency could never be
satisfied. 3.3 also calls dma_buf_set_priority(), which no mainline kernel
has; upstream's own comment says the call changes no behaviour, so it is
patched out.

Signed-off-by: Andrew Udvare <audvare@gmail.com>