Missing Kernel Patches

BlueEar writes: "There is an interesting, short story posted on the Gentoo Linux site. It talks about kernel patches created by Linux distributors that while publically available never get submitted. It even gives an example of one 'no brainer' patch that has been sitting over a year, without being incorporated into the 2.4.x distribution. The article ends with an appeal to Linux community to keep those patches flowing to Marcelo."

Comments

Issue

[mrfiddlehead]

Since the patch doesn't show how eidx has been calculated it's not immediately obvious that this patch should even be applied. That is, if the bug was subsequently "fixed" by incrementing eidx, when it was calculated, then this patch would make matters worse. So you'd have to go get the 2.4.3 source and verify that the calculation of eidx has not itself changed.

Careful.

Re:These patches can hardly be critical

[Alan Cox]

Precisely. This patch was in fact submitted and the consensus was thats its tricky to prove correct, its 1 page of memory and it was better to wait for 2.5 before doing that work.

Re:I am wondering...

[bero-rh]

I am wondering if the distributors themselves don't have too much interest in offering patches upstream

This plain isn't true, and whoever wrote the article on gentoo.org just shows he doesn't have the slightest hint of a clue.

There are some good reasons not to blindly apply distributor patches into the main kernel (for example, we have quite a few workarounds for bugs, but the right way to fix them in the official kernel is to fix them, not to add workarounds), and there are some other things preventing other patches from getting in (e.g. Linus not having the time to handle them immediately).

Other stuff is controversial (such as Red Hat Rawhide kernels putting in the Rik VM rather than the AA VM currently in upstream).

The patches are sent upstream, but at least Red Hat doesn't believe in forcing upstream maintainers to accept all patches we send.