[04:49:53] *** Joins: gila (~gila@5ED74129.cm-7-8b.dynamic.ziggo.nl) [06:38:13] *** Quits: tomzawadzki (~tomzawadz@192.55.54.42) (Ping timeout: 246 seconds) [07:50:47] *** Joins: tomzawadzki (~tomzawadz@134.134.139.72) [08:02:34] *** Quits: tomzawadzki (~tomzawadz@134.134.139.72) (Ping timeout: 246 seconds) [08:29:05] *** Joins: tomzawadzki (~tomzawadz@134.134.139.72) [08:32:43] *** Quits: tomzawadzki (~tomzawadz@134.134.139.72) (Remote host closed the connection) [09:58:22] peluse, thanks for the link to that article! Good read! [10:48:38] you bet! [10:55:47] *** Quits: klateck (~klateck@134.134.139.72) (Ping timeout: 240 seconds) [10:56:02] *** Quits: jimharris (~jimharris@134.134.139.72) (Ping timeout: 252 seconds) [10:56:18] *** Quits: jkkariu (jkkariu@nat/intel/x-cnsvagbewqfrlyzl) (Ping timeout: 264 seconds) [10:56:18] *** Quits: gangcao (gangcao@nat/intel/x-plfymhttumwsylzj) (Ping timeout: 264 seconds) [10:56:18] *** Quits: vermavis (vermavis@nat/intel/x-hmpbdrvsgwpsasmm) (Ping timeout: 264 seconds) [10:56:18] *** Quits: kjakimia (kjakimia@nat/intel/x-rggaiycmiihisuee) (Ping timeout: 264 seconds) [10:56:46] *** Quits: ppelplin (~ppelplin@134.134.139.72) (Ping timeout: 252 seconds) [10:57:08] *** Quits: pwodkowx (pwodkowx@nat/intel/x-kobwbakzkufimdpm) (Ping timeout: 252 seconds) [10:57:12] *** Quits: peluse (~peluse@134.134.139.72) (Ping timeout: 250 seconds) [10:57:30] *** Quits: sethhowe (~sethhowe@134.134.139.72) (Ping timeout: 264 seconds) [10:57:30] *** Quits: mszwed (mszwed@nat/intel/x-swaeotjlvqvequnv) (Ping timeout: 264 seconds) [11:11:12] *** Joins: jimharris (jimharris@nat/intel/x-wpdyqryiduqedtrx) [11:11:12] *** ChanServ sets mode: +o jimharris [11:21:43] *** Joins: sethhowe (~sethhowe@134.134.139.72) [11:28:44] *** Joins: pwodkowx (~pwodkowx@134.134.139.72) [11:28:51] *** Parts: pwodkowx (~pwodkowx@134.134.139.72) ("Leaving") [11:28:57] *** Joins: pwodkowx (~pwodkowx@134.134.139.72) [13:45:34] *** Joins: travis-ci (~travis-ci@ec2-54-160-151-185.compute-1.amazonaws.com) [13:45:35] (spdk/master) nvmf: Move target opts to transport opts (part 2) (John Barnard) [13:45:35] Diff URL: https://github.com/spdk/spdk/compare/f8d0f6d6ccdd...183d81d0c65d [13:45:35] *** Parts: travis-ci (~travis-ci@ec2-54-160-151-185.compute-1.amazonaws.com) () [14:26:32] *** Joins: travis-ci (~travis-ci@ec2-54-81-125-58.compute-1.amazonaws.com) [14:26:33] (spdk/master) bdev: Add crypto virtual bdev module (Paul Luse) [14:26:33] Diff URL: https://github.com/spdk/spdk/compare/183d81d0c65d...51606ed40277 [14:26:33] *** Parts: travis-ci (~travis-ci@ec2-54-81-125-58.compute-1.amazonaws.com) () [15:05:06] *** Quits: JC_Intel (86868b4c@gateway/web/freenode/ip.134.134.139.76) (Ping timeout: 252 seconds) [15:15:37] *** Joins: travis-ci (~travis-ci@ec2-23-20-213-81.compute-1.amazonaws.com) [15:15:38] (spdk/master) app: initializing log after DPDK (wuzhouhui) [15:15:38] Diff URL: https://github.com/spdk/spdk/compare/51606ed40277...9eae89f10c32 [15:15:38] *** Parts: travis-ci (~travis-ci@ec2-23-20-213-81.compute-1.amazonaws.com) () [15:16:27] *** Joins: travis-ci (~travis-ci@ec2-54-91-180-161.compute-1.amazonaws.com) [15:16:28] (spdk/master) vhost: rpc: fix memory leak of spdk_rpc_get_vhost_controllers (wuzhouhui) [15:16:28] Diff URL: https://github.com/spdk/spdk/compare/9eae89f10c32...33aa99485548 [15:16:28] *** Parts: travis-ci (~travis-ci@ec2-54-91-180-161.compute-1.amazonaws.com) () [16:22:12] *** Joins: travis-ci (~travis-ci@ec2-54-197-106-137.compute-1.amazonaws.com) [16:22:13] (spdk/master) doc: add rpc example to bdev doc for crypto (paul luse) [16:22:13] Diff URL: https://github.com/spdk/spdk/compare/33aa99485548...5968bd6242b7 [16:22:13] *** Parts: travis-ci (~travis-ci@ec2-54-197-106-137.compute-1.amazonaws.com) () [17:07:14] *** Joins: peluse (~peluse@134.134.139.72) [17:07:14] *** ChanServ sets mode: +o peluse [17:08:57] jimharris, just checked on a longer running test I started earlier. data corruption w/crypto on 64K IO 256 OIO with bdevperf that wasn't there just a week or two ago. I'll start bisecting and keep you posted, I ruled out the split code as my IO boundary size is 64K anyway but I did turn it off and re-run, still fails.... probably some silly one line cleanup somewhere, grrr [17:10:32] no worries - we can sync up in the morning and chat about whatever data you get between now and then [17:14:41] not going to be able to work on it tonight so it'll be a quick sync :) [17:15:25] oh that's fine - when you said "I'll start bisecting" I thought maybe you were going to do that now :) [17:15:45] I still have my inline trace stuff I can throw back in so should be able to get to the bottom of it reasonably quickly I hope. Just went back to the first patch w/the get_buf fix and it still fails even though I know I tested it then (famous last words) [17:15:48] how long of a test was it? [17:16:05] it's almost immediate thank god [17:16:34] gotta go play dad's uber service now... later [17:25:17] OK, now I'm really leaving but I broke in bdevperf right at the 1st mis-compare to look at the bdev_io and I don't recall seeing IO that were 127 blocks before... is this normal "(gdb) p *bdev_io [17:25:17] $1 = {bdev = 0x613000000e50, type = 1 '\001', iov = {iov_base = 0x7ffc249f8e00, iov_len = 65024}, child_iov = {{ [17:25:17] iov_base = 0x0, iov_len = 0} }, u = {bdev = {iovs = 0x7ffc2b1eaf90, iovcnt = 1, [17:25:17] num_blocks = 127" [17:26:11] what's your bdevperf command line? [17:26:52] one sec... already have this in my paste buffer. Here is the comparison and g_io_size is a full 64K "if (memcmp(task->buf, iovs[0].iov_base, g_io_size) != 0) {" [17:28:10] OMG [17:28:42] I hand types it instead of running the same test script I usually run. Instead of 65536 I hit 65535 [17:28:42] what? [17:28:48] so, works fine now :) [17:29:01] hand-typed - typed it in by hand [17:29:10] that's a bug in bdevperf then - we shouldn't allow you to pick stupid io sizes like that :) [17:29:24] yeah, this works: sudo ./bdevperf -c b.conf -q 256 -o 65536 -w verify -t 3 [17:29:33] this gives me a data corruption: sudo ./bdevperf -c b.conf -q 256 -o 65535 -w verify -t 3 [17:29:45] looks like it rounded down to the next block [17:30:01] I'll submit a patch to fix that, shit now I'm late. gotta go! thanks [17:30:31] later [19:46:35] pushed... https://review.gerrithub.io/#/c/spdk/spdk/+/425883 [23:21:24] *** Joins: klateck (klateck@nat/intel/x-maoulpntompwwiyv)