Reintroduce hfsc-based shapers

[xlfjn]

It seems that sqm-scripts used to have hfsc shapers until 39eb7a0, when it was removed due to its incompatibility with fq_codel at that time. This kernel issue has since been resolved, as recorded in a67824b.

It may be worthwhile to re-add these shapers as described in the original removal commit 39eb7a0, "This commit can be reverted at a newer kernel." It would help users test how hfsc performs in their environment.

[yutayu]

wow lol

[tohojo]

xlfjn [email protected] writes:

It seems that sqm-scripts used to have hfsc shapers until 39eb7a0, when it was removed due to its incompatibility with fq_codel at that time. This kernel issue has since been resolved, as recorded in a67824b.

It may be worthwhile to re-add these shapers as described in the original removal commit 39eb7a0, "This commit can be reverted at a newer kernel." It would help users test how hfsc performs in their environment.

Not necessarily opposed to this; but I'm not sure if it's just a matter of reverting that commit, as we've changed how we call sqm_start since those scripts were removed...

[dtaht]

Same goes also for never seeing an actual benefit to hfsc support in the first place.