[00:58:34] *** Quits: ziyeyang_ (~ziyeyang@192.55.54.44) (Quit: Leaving) [01:19:40] *** Quits: Shuhei (caf6fc61@gateway/web/freenode/ip.202.246.252.97) (Ping timeout: 260 seconds) [01:22:21] *** Joins: gila (~gila@5ED4D9C8.cm-7-5d.dynamic.ziggo.nl) [01:44:36] *** Joins: tomzawadzki (~tomzawadz@134.134.139.75) [01:50:10] *** Quits: gila (~gila@5ED4D9C8.cm-7-5d.dynamic.ziggo.nl) (Quit: My Mac Pro has gone to sleep. ZZZzzz…) [01:50:43] changpe1: why exactly do we have num-queues (with dash) rather than num_queues (with underscore) in vhost-blk in QEMU? [01:50:51] is there any reasoning behind this? [03:47:04] *** Joins: gila (~gila@5ED4D9C8.cm-7-5d.dynamic.ziggo.nl) [03:48:10] *** Quits: gila (~gila@5ED4D9C8.cm-7-5d.dynamic.ziggo.nl) (Client Quit) [03:48:32] *** Joins: gila (~gila@5ED4D9C8.cm-7-5d.dynamic.ziggo.nl) [07:59:58] darsto: saw your comment on https://review.gerrithub.io/#/c/396058/ - I thought about rolling it back too, but keeping the extra IOV there helps cover the extreme corner case where we do need to split one out of 128 IOVs - usually each VM does have two memory regions so this theoretically could happen [08:03:07] jimharris: true. a long-term solution would be to keep (max_mem_regions - 1) extra descriptors [08:03:24] max_mem_regions is 8 now, so that's doable [08:15:56] i'm thinking of a small vhost-user app that uses all 8 memory regions, 2MB each [08:27:13] *** Quits: gila (~gila@5ED4D9C8.cm-7-5d.dynamic.ziggo.nl) (Quit: My Mac Pro has gone to sleep. ZZZzzz…) [08:27:36] *** Joins: gila (~gila@5ED4D9C8.cm-7-5d.dynamic.ziggo.nl) [09:14:45] *** Quits: tomzawadzki (~tomzawadz@134.134.139.75) (Remote host closed the connection) [09:32:21] *** Quits: gila (~gila@5ED4D9C8.cm-7-5d.dynamic.ziggo.nl) (Quit: My Mac Pro has gone to sleep. ZZZzzz…) [09:41:09] darsto: nice - yeah, that sounds like a simple mechanism [09:42:47] but we've got the C1K problem coming up and ben was already lifting the 8 memory region restriction [09:43:53] so I'd leave VHOST_IOVS_MAX at 129 for now [10:35:25] *** Quits: sethhowe (sethhowe@nat/intel/x-fgmngxyagqjgusjo) (Remote host closed the connection) [10:37:01] *** Joins: sethhowe (sethhowe@nat/intel/x-ashqivebhpallqma) [11:35:16] *** Quits: sethhowe (sethhowe@nat/intel/x-ashqivebhpallqma) (Remote host closed the connection) [11:36:45] *** Joins: sethhowe (sethhowe@nat/intel/x-tejkjkniqfbwehhb) [12:11:48] *** Joins: lhodev (~Adium@inet-hqmc01-o.oracle.com) [12:34:11] *** Quits: boutcher (~boutcher@66.113.132.66) (Ping timeout: 255 seconds) [12:34:27] *** Joins: boutcher (~boutcher@66.113.132.66) [13:00:39] *** Joins: gila (~gila@5ED4D9C8.cm-7-5d.dynamic.ziggo.nl) [13:41:26] jimharris: is "blob: set md_ro and data_ro fields only after persiting data" needed for 18.01? (it needs a rebase, but I can do that if we want to get it in) [13:41:52] yes - that's needed [13:43:55] ok, rebased: https://review.gerrithub.io/#/c/394243/ [13:48:34] +2 - thanks! [13:49:20] are there any other lvol/thin provisioning patches that are ready to go and can be picked off from the series? [13:53:12] looking [13:58:43] this one is ready I think - but definitely want bwalker's feedback before we merge it [13:58:45] https://review.gerrithub.io/#/c/391422/ [14:59:45] *** Quits: gila (~gila@5ED4D9C8.cm-7-5d.dynamic.ziggo.nl) (Quit: My Mac Pro has gone to sleep. ZZZzzz…) [15:54:15] *** Joins: Shuhei (caf6fc61@gateway/web/freenode/ip.202.246.252.97) [17:02:36] *** Quits: lhodev (~Adium@inet-hqmc01-o.oracle.com) (Quit: Leaving.) [19:11:56] *** Joins: lhodev (~Adium@ip68-107-79-69.sd.sd.cox.net) [19:15:36] *** Quits: lhodev (~Adium@ip68-107-79-69.sd.sd.cox.net) (Client Quit) [19:19:44] *** Joins: lhodev (~Adium@ip68-107-79-69.sd.sd.cox.net) [19:46:27] *** Quits: lhodev (~Adium@ip68-107-79-69.sd.sd.cox.net) (Quit: Leaving.) [23:56:01] *** Joins: Aneesh (31cf3740@gateway/web/freenode/ip.49.207.55.64)