Discussion minutes from IRC #hwe (2020/06/12): 15:37:35 <tjaalton> khfeng: so, should the e1000e commits be rejected, according to your comment on the bug? 15:39:18 <khfeng> tjaalton, depends on kentlin_office. If we care about the power consumption under s0ix, then probably should reject it 15:40:02 <tjaalton> ok, I'll drop it from the list for now 15:40:27 <kentlin_office> tjaalton, khfeng, what commit 15:41:05 <khfeng> kentlin_office, the one the prevents vPro systems from entering s0ix 15:42:10 <kentlin_office> khfeng, oh.. we care about S0ix... 15:42:43 <kentlin_office> khfeng, the patch is submitted by whom? 15:44:52 <khfeng> kentlin_office, dunno 15:45:02 <-- chih (c...@1-34-63-143.hinet-ip.hinet.net) has quit (Ping timeout: 121 seconds) 15:45:29 <kentlin_office> khfeng, from upstream? 15:45:31 --> chih (c...@61-220-137-34.hinet-ip.hinet.net) has joined #hwe 15:45:55 <khfeng> kentlin_office, yea 15:45:59 <tjaalton> intel 15:47:05 <kentlin_office> tjaalton, in latest Kernel? do you have but or git cold share with me? I want to write email to Intel contact why they send such commit... 15:50:38 <tjaalton> shane: ^ 15:55:04 <tjaalton> kentlin_office: it's in linus's git tree now 15:55:11 <shane> tjaalton: got it, have talked with kentlin_office 15:55:51 <tjaalton> what's the verdict? 15:56:02 <tjaalton> should I drop both 16:06:53 <-- jeremysu (jerem...@61-220-137-37.hinet-ip.hinet.net) has quit (Ping timeout: 121 seconds) 16:07:17 --> jeremysu (jerem...@61-220-137-37.hinet-ip.hinet.net) has joined #hwe 16:08:06 <tjaalton> khfeng: I can't find your e1000e commit in the trees where it was supposed to be applied already.. do you know what happened? 16:08:19 <tjaalton> not in focal at least 16:09:28 <khfeng> tjaalton, "e1000e: Disable TSO for buffer overrun workaround" this one? 16:09:34 <tjaalton> yes 16:11:38 --> ethanhsieh_home (et...@1-171-165-81.dynamic-ip.hinet.net) has joined #hwe 16:13:00 <tjaalton> perhaps I should ask the ones who applied it 16:14:39 <tjaalton> it's in bionic at least 16:14:46 <tjaalton> but not focal. probably a mistake 16:14:54 <tjaalton> khfeng: so, nevermind :) 16:15:12 <khfeng> Ok :) 16:15:30 <-- ethanhsieh_home (et...@1-171-165-81.dynamic-ip.hinet.net) has quit (Quit: Leaving) 16:20:42 <-> alextu is now known as alextu-meeting 16:25:43 <tjaalton> vicamo: so, back to iommu.. I'm not really sure :D 16:26:12 <vicamo> tjaalton: I saw you're merging "NAK: [PATCH 0/2][SRU][B/E/F/U/OEM-OSP1-B/OEM-5.6] PCI: Avoid FLR for AMD Matisse/Starship HD Audio & USB 3.0"? 16:26:40 --> hughchao_office (hughc...@1.general.hughchao.uk.vpn) has joined #hwe 16:27:30 <tjaalton> vicamo: oh, that was a mistake.. already replied to the correct thread 16:27:54 <vicamo> khfeng: are you interested in that power button issue? are holding a conf call 16:28:04 <khfeng> vicamo, sure 16:28:06 <vicamo> tjaalton: ok, thanks 16:28:59 <-- hugh_chao (hughc...@61-220-137-37.hinet-ip.hinet.net) has quit (Ping timeout: 121 seconds) 16:31:39 <vicamo> khfeng: forwarded conf info for you 16:32:00 <tjaalton> shane: so what should I do with those patches? 16:37:18 <khfeng> vicamo, joined via phonecall? 16:37:21 <shane> tjaalton: sorry, wait a minute 16:37:39 <vicamo> khfeng: yes 16:43:40 <shane> tjaalton: Do you mean https://lists.ubuntu.com/archives/kernel-team/2020-June/110994.html ? 16:47:15 <khfeng> kentlin_office, vicamo, ... what was that 16:47:24 <khfeng> kentlin_office, purely wasting time 16:47:30 <tjaalton> shane: yes 16:50:11 <tjaalton> shane: I applied before reading the bug comment 16:51:14 <AceLan> tjaalton: the commits are from here https://bugs.launchpad.net/somerville/+bug/1872029 16:51:14 <mup> Bug #1872029: [Task] e1000e S0ix support for TGL <e1000e> <hwe-networking-ethernet> <s0ix> <tigerlake> <somerville:New for kent-jclin> <somerville hwe-somerville:In Progress by shanelin> <https://launchpad.net/bugs/1872029> 16:53:34 <tjaalton> I know it's in upstream 5.8 tree 16:54:11 <tjaalton> both of these 16:54:48 <AceLan> tjaalton: you want them to be cherry picked from master tree? 16:55:31 <tjaalton> AceLan: no, khfeng said the second patch causes issues 16:58:18 <tjaalton> shane: what did you decide after discussing with kentlin_office? 16:59:21 <shane> tjaalton: kentlin_office says he will ask AceLan.. 16:59:52 <-- kyleN (knitzs...@c-73-123-35-110.hsd1.ma.comcast.net) has quit (Ping timeout: 121 seconds) 17:01:27 <AceLan> humm 17:02:00 <tjaalton> haha 17:02:34 --> kyleN (knitzs...@c-73-123-35-110.hsd1.ma.comcast.net) has joined #hwe 17:02:44 <tjaalton> and should I just apply that iommu patchset? guess we'll find out soon enough it the world blows up.. 17:03:20 <AceLan> khfeng: ^^ # from the e1000e commit description it says "Since ME systems do not support SLP_S0 in S0ix state, and S0ix entry and exit flows may cause errors on them" do you still think we shouldn't apply this commit? 17:04:11 <kentlin_office> AceLan, so we need this patch to enter SLP_S0? 17:04:26 <khfeng> AceLan, depends on if we want the systems reach s0ix 17:04:30 <kentlin_office> AceLan, base on what commit say? 17:04:38 <khfeng> kentlin_office, the patch _prevents_ s0ix 17:04:55 <AceLan> kentlin_office: https://lists.ubuntu.com/archives/kernel-team/2020-June/110805.html 17:06:54 <kentlin_office> AceLan, khfeng so only e1000e could not enter s0ix with this patch.. is that what it meant or the whole system? 17:07:45 <khfeng> kentlin_office, ?????? 17:08:32 <AceLan> kentlin_office: only the prevent the listed nic i217, i218, and i219 from entering s0ix 17:10:40 <khfeng> kentlin_office, s0ix is an SoC state... 17:11:11 <AceLan> khfeng: kentlin_office: I think we should have this commit and once we encounter the power consumption issue we can call out for help from intel 17:12:33 <khfeng> AceLan, there's a heated discussion between Lenovo and Intel, and the current conclustion is that they won't help :) 17:13:05 <khfeng> AceLan, you can ask AaronMa on the e1000e issue... 17:13:41 <AceLan> khfeng: but the commit claims it leads to issues if those chips enter s0ix 17:14:18 <khfeng> AceLan, the patch from AaronMa can workaround the issue, however e1000e developers don't like the approach 17:15:41 <AceLan> khfeng: https://lkml.org/lkml/2020/3/23/809 # this one? 17:15:59 <khfeng> AceLan, yea that's the one 17:16:15 <kentlin_office> AceLan, is it in the mainline Kernel or Linus Kernel tree? or it is still in Intel next or somewhere else 17:16:57 <AceLan> kentlin_office: it's in v5.8 17:17:44 <AceLan> khfeng: does AaronMa backport his commit to ubuntu kernel? 17:18:08 <AaronMa> Acelan yes 17:18:49 <AaronMa> Acelan it is not in any tree because of Intel. 17:19:34 <AaronMa> Acelan only in Ubuntu kernel tree, you should apply it to oem kernel 17:21:19 <AceLan> AaronMa: to which series of kernel do you submit? 17:22:56 <AaronMa> AceLan: [OEM-OSP-1/E/F/U/OEM-5.6] 17:24:09 <khfeng> kentlin_office, "s0ix is not POR for Linux" 17:24:15 <khfeng> kentlin_office, that's their answer 17:24:19 <khfeng> kentlin_office, don't bother to ask 17:25:34 <AceLan> AaronMa: got it 17:26:12 <AceLan> khfeng: kentlin_office: okay, in this case we don't need the commit to prevent NICs from entering s0ix 17:27:17 <AceLan> agree? 17:29:09 <-- jeremysu (jerem...@61-220-137-37.hinet-ip.hinet.net) has quit (Ping timeout: 121 seconds) 17:31:30 --> jeremysu (jerem...@61-220-137-37.hinet-ip.hinet.net) has joined #hwe 17:33:00 <AceLan> sounds like no opposing opinion 17:33:36 <AceLan> tjaalton: it looks like we can drop the seconds e1000e patch, since we have a fix for the s0ix issue 17:37:13 <-- haosheng (haosh...@61-220-137-37.hinet-ip.hinet.net) has quit (Ping timeout: 121 seconds) 17:41:41 <-> alextu-meeting is now known as alextu 17:46:45 <AaronMa> AceLan: tjaalton https://lists.ubuntu.com/archives/kernel-team/2020-April/108691.html 17:47:15 <AaronMa> AceLan: tjaalton this patch is not in oem-osp-1 or oem-5.6 17:48:17 <tjaalton> AceLan, AaronMa: cool, I'll do that and add the missing one 17:49:33 <-- ivanhu (iva...@112-104-28-97.adsl.dynamic.seed.net.tw) has quit (Quit: Leaving) 17:53:01 <kentlin_office> khfeng, Then I want that in email than use that to tell Dell ha ha... 17:56:01 --> PHLin (po-...@1.general.phlin.uk.vpn) has joined #hwe 18:00:52 <tjaalton> AaronMa: actually the patch is in oem-5.6, somehow failed to report that 18:03:05 <AaronMa> tjaalton: oh, yes, in oem-5.6 branch. 18:03:17 <tjaalton> ah right, it came from the unstable tree
-- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1881996 Title: Enable e1000e S0ix on Tiger Lake To manage notifications about this bug go to: https://bugs.launchpad.net/hwe-next/+bug/1881996/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs