The Linux Desktop Responsiveness Patches Are Feeling Good

[phoronix]

Phoronix: The Linux Desktop Responsiveness Patches Are Feeling Good

As was reported on Phoronix yesterday, the Linux desktop responsiveness problem may be fixed. This is the issue that has affected many Linux desktop users for numerous months where when dealing with large file transfers or other disk operations, the desktop interface (regardless of whether its GNOME, KDE, Xfce, etc) would become unresponsive and it could be a good number of seconds before a simple action like clicking a menu item would be processed...

http://www.phoronix.com/vr.php?view=ODQ3OQ

[Shining Arcanine]

Your assumption is wrong. KernelofTruth's backport only contains 2 out of 7 of the patches. There are 5 additional that he did not backport:

http://forums.gentoo.org/viewtopic-p...0.html#6377520

Please try to do fact checks on your assumptions in the future before reporting them.

[DeepDayze]

Oh there's 7 patches that all together fix the issue?

[kernelOfTruth]

Oh there's 7 patches that all together fix the issue?

indeed, but they are more tricky to backpart

at least starting with the 3rd patch I tried it and parts of the code are spread all over the file

since I'm not that experienced I'm waiting for someone else to backport (most favorably the zen-kernel devs )

[kernelOfTruth]

since I'm pretty busy - I unfortunately can't do the further backporting

anyways: the results already are pretty impressive considering that it's only a small amount of code that's been added / removed (those 2 patches)

[kebabbert]

I thought this was FUD, but it was true? Linux could not handle things without lagging sometimes? And now, is the situation better or does it still occur?

[kernelOfTruth]

I thought this was FUD, but it was true? Linux could not handle things without lagging sometimes? And now, is the situation better or does it still occur?

it was true for some time in the past (for me)

from my observation it most probably got worse / introduced after 2.6.34 since that kernel was working excellent for me back then - even under heavy traffic


luckily with those lots of changes and the other improvements coming (e.g. reducing barrier writing, a unified slab (v3), and more) the future looks bright

[kernelOfTruth]

2.6.22 (?) with Con's SD or RSD cpu scheduler was another masterpiece in terms of performance & responsibility

[ChrisXY]

I thought this was FUD, but it was true? Linux could not handle things without lagging sometimes? And now, is the situation better or does it still occur?

Sometimes it happened and sometimes it didn't.
For example yesterday I put a 2.8 gb file on my local lighttp and downloaded it with aria2. This has to be the most I/O one would normally produce and my desktop was perfectely responsive...

[DeepDayze]

indeed, but they are more tricky to backpart

at least starting with the 3rd patch I tried it and parts of the code are spread all over the file

since I'm not that experienced I'm waiting for someone else to backport (most favorably the zen-kernel devs )

Damentz (he's one of the zen devs) has some experience with backporting patches so he might take a crack at backporting all 7 patches to 2.6.35 (and maybe 2.6.34)