)]}' { "commit": "961d0e5b32946703125964f9f5b6321d60f4d706", "tree": "fcb396f0cf79c5b519fd171879e67ce09393eabf", "parents": [ "81573b18f26defe672a7d960f9af9ac2c97f324d" ], "author": { "name": "Zh-yuan Ye", "email": "ye.zh-yuan@socionext.com", "time": "Tue Mar 24 17:28:25 2020 +0900" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Mar 24 16:14:05 2020 -0700" }, "message": "net: cbs: Fix software cbs to consider packet sending time\n\nCurrently the software CBS does not consider the packet sending time\nwhen depleting the credits. It caused the throughput to be\nIdleslope[kbps] * (Port transmit rate[kbps] / |Sendslope[kbps]|) where\nIdleslope * (Port transmit rate / (Idleslope + |Sendslope|)) \u003d Idleslope\nis expected. In order to fix the issue above, this patch takes the time\nwhen the packet sending completes into account by moving the anchor time\nvariable \"last\" ahead to the send completion time upon transmission and\nadding wait when the next dequeue request comes before the send\ncompletion time of the previous packet.\n\nchangelog:\nV2-\u003eV3:\n - remove unnecessary whitespace cleanup\n - add the checks if port_rate is 0 before division\n\nV1-\u003eV2:\n - combine variable \"send_completed\" into \"last\"\n - add the comment for estimate of the packet sending\n\nFixes: 585d763af09c (\"net/sched: Introduce Credit Based Shaper (CBS) qdisc\")\nSigned-off-by: Zh-yuan Ye \u003cye.zh-yuan@socionext.com\u003e\nReviewed-by: Vinicius Costa Gomes \u003cvinicius.gomes@intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n", "tree_diff": [ { "type": "modify", "old_id": "b2905b03a432f8fb46f06d939c847926171f79c3", "old_mode": 33188, "old_path": "net/sched/sch_cbs.c", "new_id": "2eaac2ff380fafaf9dbaeade2b86e029bddceecc", "new_mode": 33188, "new_path": "net/sched/sch_cbs.c" } ] }