Digging in the kernel dust

Digging in the kernel dust

A very enjoyable read, it's always great to hear what develops out of the RT work! Those interested in seeing Thomas's talk can do so here.

Sadly, a week before Kernel Recipes 2017, and more than a year after the rework patches were merged, a user-space-visible regression surfaced. This issue is ongoing, because members of the network community (who were most affected by this regression) initially responded by arguing that the right fix was the removal of the whole timer wheel rework — until it was pointed out that the network subsystem would suffer the biggest performance hit from doing so.

I was curious about the above issue, so went digging in the mailing list dust for the original threads:

[PATCH RESEND 0/2] enable hires timer to timeout datagram socket

[PATCH v2 0/2] enable hires timer to timeout datagram socket

[In future articles it would be great if you could include links to the talk(s) and mailing list discussion(s), to make it easy for interested readers to get the details.]

It's been almost a month since Thomas's talk, and as far as I can tell discussion about the above issue died after Thomas's last email of 25 September. No new mailing list threads on the subject have appeared (at least, that I could find) and the only change to /net/core/datagram.c since then is unrelated.