Arjan van de Ven
Arjan van de Ven
adding kernel command lines options that persist is actually easy and documented in the man page yes it's somewhat different than other distros but quite consistent and elegant On Sat,...
one more step, run clr-boot-manager update but otherwise yes On Sat, Sep 3, 2022, 16:21 Gabriel Al-Ghalith ***@***.***> wrote: > Well, you might be right, if this is what you're...
Release: is the build sequence number, and cannot be based on some upstream version thing... I'm fixing the package.
I wonder if we do not include the initrd stuff in the garbage collection algorithm though On Thu, Jun 13, 2019 at 10:34 PM ahkok wrote: > This is normal....
we actually recently packaged 18, but a "openjdk", so without the version label... On Fri, May 20, 2022 at 4:17 AM Thomas Skjølberg ***@***.***> wrote: > Hi, > > I...
also, 22 is .. unfortunate. https://github.com/xianyi/OpenBLAS/pull/1846 which got merged as https://github.com/xianyi/OpenBLAS/commit/f1c02273cb9206a43968a0d813d1e9506612343c is what I did on skylake to avoid some really bad behavior; one of these days I need to...
maybe. One might need to do the equivalent of https://github.com/xianyi/OpenBLAS/pull/1846/commits/dcc5d6291e7b02761acfb6161c04ba1f8f25b502 to avoid nasty "we did not think a work chunk could be 0 sized" behavior
before the rounding to preferred sizes it wouldn't. but lets say width is 64, and 3 cpus. preferred size is 32 before you get 21 21 22 as sizes with...
it goes further out than that, up to M * N * K = 4 * 512 * 512 (outside of that there's still wins but also some losses)
it will be correct, but the optimization goes further out than that. the MNK multiple is the best test I've found so far (I created a few thousand random M/N/K...