The Disappointing Direction Of Linux Performance From 4.16 To 5.4 Kernels

[name99]

Those are probably not regression but mitigations hits and those won't go away until those problem are fixed in silicon in the future, in fact expect it to get even worse from now on since every time a new one is discovered it seems to be way worse than the previous ones, specially for Intel CPUs.

You were saying?... Today's update

https://www.intel.com/content/www/us/en/support/articles/000055650/processors/intel-xeon-processors.html

0-4% performance hit for the microcode fix.

[starshipeleven]

AMD hardware dissipates too much power.

This has been correct only for one or two GPU generations back in 2014-15 I think. NVIDIA also had power hungry GPU generations in the past.

Would you mind not posting this as a fact that is true now?

[starshipeleven]

Please elaborate (do also note that birdie used Apache and MySQL as examples, the 40% decrease that he talks about [that would require a 66% increase in performance to compensate btw] is for system-call heavy applications which will include every single database and web server in existence).

Given that a lot of websites are commonly developed with complete shit frameworks that make them haxxor magnets, or not very stable for the very least, it is common practice to keep them on separate systems or VMs, so you have a database server and a webserver.

[starshipeleven]

Nvidia has supported Linux for far longer than AMD has.

Incorrect, NVIDIA has provided a blob driver for Linux for far longer.
Proper Linux support is opensource drivers.

[King Mucus]

Ext4 on my Box is completely borked with 5.4! RootFs gets damaged while trying to mount:

EXT4-fs error (device nvme0n1p5): ext4_journal_check_start:61: Detected aborted journal
EXT4-fs (nvme0n1p5): Remounting filesystem read-only
EXT4-fs (nvme0n1p5): ext4_writepages: jbd2_start: 8191 pages, ino 262253; err -30

That's some laudable effort by your FOSS environment to be true to the principles of freedom and be duly liberating to females, transgendered persons, religious minorities and people of colour. As you might be aware, The author of Ext4 has been accused by credible proponents of The Movement and The People as being a despicable rape apologist.

I think you should be banned from these forums for even daring contemplating the usage of such an evil, horrible and not people-oriented piece of software.

[Paul Frederick]

Incorrect, NVIDIA has provided a blob driver for Linux for far longer.
Proper Linux support is opensource drivers.

In case you are unaware Nvidia licenses technology from others. It is not theirs to just give away either. So a binary driver is the absolute best they can do. 70 years beyond the death of an author that may change? Until then Nvidia is bound by copyright law just as much as anyone else is.

[betam4x]

Am I reading that correctly? atime is enabled?

[starshipeleven]

So a binary driver is the absolute best they can do.

A binary driver still isn't "proper support" for a Linux system. Opensource mainline drivers are.

And I don't give a shit about your excuses, Intel, AMD, IBM, Broadcom and others have proper open drivers. Why NVIDIA is different?

Ah, because they don't care to do a good job at it, blobs are enough for pleb fanbois after all.

[betam4x]

Wow after reading all the comments here, it has become clear that there are quite a few ignorant users here.

This is coming from someone, who, as an ignorant teenager many years ago, was a black hatter.

Point 1: If any of the CPU exploits are in use, you won't know it. Someone that figures out how to utilize any of the exploits isn't going to say anything.

Point 2: It is POSSIBLE to write specially crafted javascript that can take advantage of a stealth exploit that does a complete stealth takeover of your machine. Depending on the exploit, it is then possible to do things like modify the bios to load exploited code, etc.

I am not saying it has been done, but it is most certainly POSSIBLE. That fact alone should give you pause. You folks disabling mitigations and other security features are nuts.

Hard to exploit vulnerabilities that have been exploited stay private. Quite often, only your data or access to your machine is sold or used fore nefarious purposes, not the vulnerability itself.

Point 3: You can't detect the code doing the exploit because of the nature of the vulnerabilities. When they are exploited (if they haven't been already), the general public won't find out until someone makes a mistake.

[geearf]

In case you are unaware Nvidia licenses technology from others. It is not theirs to just give away either. So a binary driver is the absolute best they can do. 70 years beyond the death of an author that may change? Until then Nvidia is bound by copyright law just as much as anyone else is.

That's a poor excuse, if others GPU companies managed, especially smaller ones like AMD, nVidia could too if it wanted.
I'd also guess most of these to-hide-technologies would no be in the kernel driver.
Finding their drivers/hardware better sure, arguing about it all day ok, but I fail to see the point of trying to defend nVidia on the topic of not opening their driver.