v4.19.52: patch and config update for v4.19.52 on AArch64 by Pennyzct · Pull Request #592 · kata-containers/packaging

Pennyzct · 2019-06-20T10:32:39Z

Hi~ @egernst I have already done patch and config update for v4.19.52 on AArch64.

Pennyzct · 2019-06-20T10:33:49Z

@gnawux @egernst soooooo desperately looking forward to bump to v5.0.x. ;)

grahamwhaley

presuming HOTPLUG additions are deliberate,
lgtm

grahamwhaley · 2019-06-20T10:37:22Z

kernel/configs/arm64_kata_kvm_4.19.x

 # CONFIG_ACPI_DEBUG is not set
 # CONFIG_ACPI_PCI_SLOT is not set
 CONFIG_ACPI_CONTAINER=y
+CONFIG_ACPI_HOTPLUG_MEMORY=y


I presume the HOTPLUG additions are deliberate. May want to add a note to the commit message, as they look like not just part of an update, but are an addition.

yep, you just got me😱. I may need to split that to another commit...🤠

There are several entries added in this patch. Do you have any explanation here @Pennyzct?

gnawux · 2019-06-20T10:37:51Z

I could understand your feeling 🤔

I prefer to stick on 4.19 before next LTS kernel is released, however, I don't mind aarch64 have a different kernel version.

Pennyzct · 2019-06-20T10:43:39Z

@gnawux just imaging keep refining this whole big patch. it's killing me. dose everyone on board for AArch64 has different kernel version?? @egernst @jodh-intel @devimc @grahamwhaley

gnawux · 2019-06-20T10:48:50Z

@Pennyzct

dose everyone on board for AArch64 has different kernel version??

I think this idea is a double-sword for you. It may help you avoid pick huge patches from non-LTS kernels, on the other hand, it might make others less care of ARM arch by some means.

I can't tell if it is a better choice for you.

gnawux · 2019-06-20T10:50:31Z

@grahamwhaley could we set this PR as a dependency of CI of kata-containers/runtime#1817?

Pennyzct · 2019-06-20T10:52:08Z

@gnawux emmmmmmm, But I have memory hot-remove, DAX, cpu hotplug, kvm-ptp on the way. 😱🙃

grahamwhaley · 2019-06-20T10:54:34Z

@gnawux @egernst soooooo desperately looking forward to bump to v5.0.x. ;)

Well, there was a PR for that kata-containers/runtime#1504 , but it got closed very recently due to 'inactivity', which maybe was not the right course of action.
It got stalled as the kernel fragment PR got stuck on:

it seemed to make the firecracker CI less stable
it really should have SRIOV support added in

Personally, I suspect the right path is to move to v5.x kernels with full config files, and carry on with the fragment support work in parallel - unless there is a focussed effort to land the fragment work rsn.

gnawux · 2019-06-20T10:57:17Z

emmmmmmm, But I have memory hot-remove, DAX, cpu hotplug, kvm-ptp on the way. 😱🙃

@Pennyzct awesome, how about co-present a topic on OpenInfra Summit Shanghai this November on your excellent upstream work.

jcvenegas · 2019-06-20T16:27:12Z

/test

jcvenegas · 2019-06-20T16:28:36Z

@Pennyzct could you add a depens on kata-containers/runtime#1817

jcvenegas · 2019-06-20T16:29:27Z

@egernst could you add a depends-on #592 ?

Update to newer stable kernel Fixes: kata-containers#1816 Depends-on: kata-containers/packaging#592 Signed-off-by: Eric Ernst <eric.ernst@intel.com>

Update to newer stable kernel Fixes: kata-containers#1816 Depends-on: github.com/kata-containers/packaging#592 Signed-off-by: Eric Ernst <eric.ernst@intel.com>

Pennyzct · 2019-06-21T03:09:41Z

Hi~ @bergwolf @jcvenegas I have already added depends-on. ;)

gnawux · 2019-06-21T03:16:15Z

/test

gnawux · 2019-06-21T04:32:27Z

/retest

gnawux · 2019-06-21T04:39:29Z

checked the failure in the ARM test of kata-containers/runtime#1817, it is caused by lacking ptp0 device, however, looks the fix in kata-containers/osbuilder#265 won't be merged soon.

Let's merge this and kata-containers/runtime#1817 once the essential CIs are passed because this is a security related fix.

gnawux

Thanks @Pennyzct for rebasing the patch

bergwolf · 2019-06-21T06:03:30Z

CI failed at ERROR: Please bump version in kernel/kata_config_version. @Pennyzct please bump kata_config_version.

Pennyzct · 2019-06-21T07:14:11Z

Hi~ @bergwolf while you are on children-happy hour, I have done the update. ;)

Pennyzct · 2019-06-21T07:16:22Z

Hi~ @grahamwhaley I'm too lazy to split a new one, I'm guilty😈, I just address HOTPLUG additions on this commit

gnawux · 2019-06-21T07:20:59Z

/test

gnawux · 2019-06-21T08:23:57Z

again

ERROR: Please bump version in kernel/kata_config_version

we need to do patch and config update for v4.19.52 on AArch64. The config file adds a few configs involved with memory hot-plug support. Fixes: kata-containers#591 Depends-on: github.com/kata-containers/runtime#1817 Signed-off-by: Penny Zheng <penny.zheng@arm.com>

Pennyzct · 2019-06-21T08:50:13Z

/test

gnawux · 2019-06-21T11:00:38Z

finally passed, merge it

Update to newer stable kernel Fixes: kata-containers#1816 Depends-on: github.com/kata-containers/packaging#592 Signed-off-by: Eric Ernst <eric.ernst@intel.com>

bergwolf · 2019-06-21T12:17:23Z

Thanks @Pennyzct!

devimc · 2019-06-21T12:52:41Z

@gnawux snap-ci was failing and you didn't care about it and merged this anyway!!! 👎

I can't release the new snap because it's broken, Thanks @gnawux nice job!! 👎

jodh-intel · 2019-06-21T13:06:08Z

@devimc, @gnawux - Relax guys 😎 Mistakes happen. Nobody got hurt right?

I think we need to all focus on fixing the CI's so that a failure is a very unusual occurrence (/me looks sideways at kata-containers/osbuilder#265 as that is causing constant failures .... ;)

But for now, can we take a deep breath, count to 10 and think about what we can do to fix this?

jodh-intel · 2019-06-21T13:07:32Z

@Pennyzct - do you have any calming emojis you can drop over here? 😉

gnawux · 2019-06-22T07:22:30Z

@devimc,

Penny explained the snap-ci situation for me -- the snap-ci download the old kernel and will always fail for the PR. On the other hand, the 4.19.52 is a security update for a couple of CVEs, we can't wait for fixing all CIs. For a security update, we should merge it once it got enough +1 and all the required CI passed.

On the other hand, one +1 is from me. My thoughts on it: I don't think this patch broke any existing stuff

Firstly, this is an arch specific update, and it won't affect other archs.
Secondly, the aarch64 developer Penny confirmed it worked and didn't break more cases on either packaging tests and runtime CIs.

Finally, as James stated, the only failed aarch64 case in runtime repo is the chrony one, we should solve that soon.

Have a good weekend.

Pennyzct · 2019-06-24T04:40:09Z

Hi~
@jodh-intel 🍨🍧🍦 ice-cream is always the best option here.
the fault comes from my huge, obnoxious kernel patches😡🤬, people are asking for 4.19.52, which is a security update for a couple of CVEs. So i raised this PR to do a little, tiny twist for the changing from 4.19.28 to 4.19.52.
@devimc FWIT, kernel version for snap stays in v1.8.0-alpha0, which is still the old 4.19.28. Soooooo, o(TヘTo)
I think we could include eric's kernel update PR runtime/1817 into a kata new version, maybe v1.8.0-alpha4. and update snap to this version v1.8.0-alpha4, and it will get fixed. ;)

jodh-intel · 2019-06-24T07:38:14Z

Thanks @Pennyzct - 🍨 sounds great, particularly with 🍒 and 🍫 on top ;)

Let's also add 🎵, a ⛲, 🌞, 😎 and some relaxing 🐟 to watch 😄

Your plan sounds good to me - wdyt @egernst?

gnawux · 2019-06-26T14:34:42Z

@bergwolf Do as what @Pennyzct said?

I think we could include eric's kernel update PR runtime/1817 into a kata new version, maybe v1.8.0-alpha4. and update snap to this version v1.8.0-alpha4, and it will get fixed. ;)

bergwolf · 2019-06-27T01:51:17Z

@gnawux Yeah, now that all the kernel update PRs are merged, let's roll out a series of new releases.

grahamwhaley approved these changes Jun 20, 2019

View reviewed changes

gnawux mentioned this pull request Jun 20, 2019

versions update kernel to 4.19.52 kata-containers/runtime#1817

Merged

egernst pushed a commit to egernst/runtime that referenced this pull request Jun 20, 2019

versions update kernel to 4.19.52

2500269

Update to newer stable kernel Fixes: kata-containers#1816 Depends-on: kata-containers/packaging#592 Signed-off-by: Eric Ernst <eric.ernst@intel.com>

egernst pushed a commit to egernst/runtime that referenced this pull request Jun 20, 2019

versions update kernel to 4.19.52

8637f73

Update to newer stable kernel Fixes: kata-containers#1816 Depends-on: github.com/kata-containers/packaging#592 Signed-off-by: Eric Ernst <eric.ernst@intel.com>

egernst pushed a commit to egernst/runtime that referenced this pull request Jun 21, 2019

versions: update kernel to 4.19.52

749dd04

Update to newer stable kernel Fixes: kata-containers#1816 Depends-on: github.com/kata-containers/packaging#592 Signed-off-by: Eric Ernst <eric.ernst@intel.com>

Pennyzct force-pushed the v4.19.52 branch from e1d2259 to cf4f1e5 Compare June 21, 2019 03:08

gnawux approved these changes Jun 21, 2019

View reviewed changes

Pennyzct force-pushed the v4.19.52 branch from cf4f1e5 to aacad0e Compare June 21, 2019 07:12

Pennyzct force-pushed the v4.19.52 branch from aacad0e to 77ef7e3 Compare June 21, 2019 08:49

gnawux merged commit fdacac9 into kata-containers:master Jun 21, 2019

This was referenced Jun 21, 2019

versions: update kernel to 4.19.52 kata-containers/runtime#1819

Merged

versions: update kernel to 4.19.52 kata-containers/runtime#1820

Merged

Pennyzct mentioned this pull request Jul 5, 2019

nemu: enable compile nemu for arm64 on x86 host #578

Merged

Conversation

Pennyzct commented Jun 20, 2019

Uh oh!

Pennyzct commented Jun 20, 2019

Uh oh!

grahamwhaley left a comment

Choose a reason for hiding this comment

Uh oh!

grahamwhaley Jun 20, 2019

Choose a reason for hiding this comment

Uh oh!

Pennyzct Jun 20, 2019

Choose a reason for hiding this comment

Uh oh!

gnawux Jun 20, 2019

Choose a reason for hiding this comment

Uh oh!

gnawux commented Jun 20, 2019

Uh oh!

Pennyzct commented Jun 20, 2019

Uh oh!

gnawux commented Jun 20, 2019

Uh oh!

gnawux commented Jun 20, 2019

Uh oh!

Pennyzct commented Jun 20, 2019

Uh oh!

grahamwhaley commented Jun 20, 2019

Uh oh!

gnawux commented Jun 20, 2019

Uh oh!

jcvenegas commented Jun 20, 2019

Uh oh!

jcvenegas commented Jun 20, 2019

Uh oh!

jcvenegas commented Jun 20, 2019

Uh oh!

Pennyzct commented Jun 21, 2019

Uh oh!

gnawux commented Jun 21, 2019

Uh oh!

gnawux commented Jun 21, 2019

Uh oh!

gnawux commented Jun 21, 2019

Uh oh!

gnawux left a comment

Choose a reason for hiding this comment

Uh oh!

bergwolf commented Jun 21, 2019

Uh oh!

Pennyzct commented Jun 21, 2019

Uh oh!

Pennyzct commented Jun 21, 2019

Uh oh!

gnawux commented Jun 21, 2019

Uh oh!

gnawux commented Jun 21, 2019

Uh oh!

Pennyzct commented Jun 21, 2019

Uh oh!

gnawux commented Jun 21, 2019

Uh oh!

bergwolf commented Jun 21, 2019

Uh oh!

devimc commented Jun 21, 2019

Uh oh!

jodh-intel commented Jun 21, 2019

Uh oh!

jodh-intel commented Jun 21, 2019

Uh oh!

gnawux commented Jun 22, 2019

Uh oh!

Pennyzct commented Jun 24, 2019

Uh oh!

jodh-intel commented Jun 24, 2019

Uh oh!

gnawux commented Jun 26, 2019

Uh oh!

bergwolf commented Jun 27, 2019

Uh oh!