Linux 5.11-rc1 Released - Many New Features While Dominated By AMD Header Additions

[perpetually high]

I don't use VirtualBox, but I did a quick test with Arch in a container. virtualbox-host-dkms-6.1.16-4 builds just fine with both stock kernel and linux-5.10.3.arch1-1.

Not sure, I'm using a newer build than that (6.1.17) and no luck. Hoping it'll get worked out soon.

$ sudo ./VirtualBox-6.1.17-141968-Linux_amd64.run
Verifying archive integrity... All good.
Uncompressing VirtualBox for Linux installation.............
VirtualBox Version 6.1.17 r141968 (2020-12-18T22:03:34Z) installer
Installing VirtualBox to /opt/VirtualBox
Python found: python, installing bindings...
vboxdrv.sh: failed: Look at /var/log/vbox-setup.log to find out what went wrong.

There were problems setting up VirtualBox. To re-start the set-up process, run
/sbin/vboxconfig
as root. If your system is using EFI Secure Boot you may need to sign the
kernel modules (vboxdrv, vboxnetflt, vboxnetadp, vboxpci) before you can load
them. Please see your Linux system's documentation for more information.

VirtualBox has been installed successfully.

[caligula]

There still isn't a working VirtualBox build compatible with 5.10, and now we're at 5.11-rc1. I keep a recently compiled 5.4 LTS kernel around for this reason, since 5.9 is now EOL. First world problems I guess?

I don't use VirtualBox, but I did a quick test with Arch in a container. virtualbox-host-dkms-6.1.16-4 builds just fine with both stock kernel and linux-5.10.3.arch1-1.

[perpetually high]

There still isn't a working VirtualBox build compatible with 5.10, and now we're at 5.11-rc1. I keep a recently compiled 5.4 LTS kernel around for this reason, since 5.9 is now EOL. First world problems I guess?

[skeevy420]

Interesting... when I try to edit my own posts I now get the following:

Error

Error information: " 403 error Forbidden"

That was happening to me on Christmas Day in that Fedora Zine thread so I used GIMP to edit my recipoem into a text filled picture and that worked. It happened anytime I'd click edit or preview on that thread.

[cl333r]

Linux 5.11rc1 failed to build on the kernel ppa:

https://kernel.ubuntu.com/~kernel-ppa/mainline/v5.11-rc1/

[f0rmat]

So all these regressions with AMD processors have been fixed?

From the article...

"Linux 5.11 stable should be out in February. I've already been working on many Linux 5.11 kernel benchmarks and things are looking good besides the AMD performance regression with the Schedutil governor for Zen 2 and newer where frequency invariance data is now used... More tests and insight there in the next day or two as been hammering several systems with benchmarks and different configurations to try to improve the functionality."

So the article seems to imply that at best, unknown, and at worse, no. Who knows, maybe Michael during his benchmarking and configuration changes will either find a work around or possible find the cause.

[piorunz]

So all these regressions with AMD processors have been fixed?

[bridgman]

Interesting... when I try to edit my own posts I now get the following:

Error

Error information: " 403 error Forbidden"

[bridgman]

Out of curiosity is it a case that once these register entries are compiled in they are actually small in the kernel or is half of the compiled kernel register settings for various devices?

The register header files don't take up any space once compiled, except to the extent that entries are actually used.

Put differently, if we trimmed the header files down so they only contained the registers and fields that were used by the current driver code there would be zero difference in the size of the compiled driver.

[MadeUpName]

As I am looking towards building a new HTPC when Van Gough is released this is good news. A YT talking head mentioned that some one poking around in the AMD commits had found an entry for a new threadripped part based on the new generation. I'm guessing that will be the next major dump. I'm sure Michael would not let some thing like that slide.

Out of curiosity is it a case that once these register entries are compiled in they are actually small in the kernel or is half of the compiled kernel register settings for various devices?