Skip to contentSkip to conversation

Google go CVEs

kirkstone Valek, Andrej 6 participants 12 messages Mar 7, 2023

lore
Share & tools
lore
#0 lore raw

Hello everyone,

I would like to ask you how to proceed with multiple CVEs for Google Go component in kirkstone branch.

CVEs in current version 1.17.13:
- CVE-2022-41722
- CVE-2022-41725
- CVE-2022-41724
- CVE-2022-41723

They are fixed in 1.19.6/1.20.1 branches, but a fixing patches are available for all of them too. Unfortunately there is more then ~1000 changed LOC. So not sure if this is the right approach to apply them. Not sure if the upgrade is acceptable.

So how to proceed with this?

I know, that they aren't a critical one, but would be nice to have them fixed.

Regards,
Andrej

#1 lore raw

You probably should make a kirkstone mixin layer like we did for dunfell.
https://git.yoctoproject.org/meta-lts-mixins/

Alex

Quoted 33 lines · Andrej Valek

Hello everyone,

I would like to ask you how to proceed with multiple CVEs for Google Go component in kirkstone branch.

CVEs in current version 1.17.13:
- CVE-2022-41722
- CVE-2022-41725
- CVE-2022-41724
- CVE-2022-41723

They are fixed in 1.19.6/1.20.1 branches, but a fixing patches are available for all of them too. Unfortunately there is more then ~1000 changed LOC. So not sure if this is the right approach to apply them. Not sure if the upgrade is acceptable.

So how to proceed with this?

I know, that they aren't a critical one, but would be nice to have them fixed.

Regards,
Andrej

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#178089): https://lists.openembedded.org/g/openembedded-core/message/178089
Mute This Topic: https://lists.openembedded.org/mt/97444547/1686489
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

#2 lore raw

Hello Alex,

Yes, that would an option, but afaik it wasn't working quite well. So I would still prefer a straight forward solution.

Should I spend some time for creating such patches? Means if there will be a potential option for being accepted?

Andrej

Quoted 48 lines · Alexander Kanavin

You probably should make a kirkstone mixin layer like we did for dunfell.
https://git.yoctoproject.org/meta-lts-mixins/

Alex

Quoted 40 lines · Andrej Valek

Hello everyone,

I would like to ask you how to proceed with multiple CVEs for Google Go
component in kirkstone branch.

CVEs in current version 1.17.13:
- CVE-2022-41722
- CVE-2022-41725
- CVE-2022-41724
- CVE-2022-41723

They are fixed in 1.19.6/1.20.1 branches, but a fixing patches are available for all of them too. Unfortunately there is more then ~1000
changed LOC. So not sure if this is the right approach to apply them.
Not sure if the upgrade is acceptable.

So how to proceed with this?

I know, that they aren't a critical one, but would be nice to have them
fixed.

Regards,
Andrej

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group. View/Reply Online (#178089):
https://lists.openembedded.org/g/openembedded-core/message/178089
Mute This Topic: https://lists.openembedded.org/mt/97444547/1686489
Group Owner: [email protected] Unsubscribe:
https://lists.openembedded.org/g/openembedded-core/unsub [
[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

#3 lore raw

If you understand the code well, and can be confident that your backports address the issue correctly and do not introduce new issues, then by all means go ahead.

My personal position should be known: I see the whole 'CVE backporting' industry as a colossal waste. We need to learn to update to supported upstream versions, and not be scared of breaking production with that.

Alex

Quoted 61 lines · Andrej

Hello Alex,

Yes, that would an option, but afaik it wasn't working quite well. So I would still prefer a straight forward solution.

Should I spend some time for creating such patches? Means if there will be a potential option for being accepted?

Andrej

Quoted 48 lines · Alexander Kanavin

You probably should make a kirkstone mixin layer like we did for dunfell.
https://git.yoctoproject.org/meta-lts-mixins/

Alex

Quoted 40 lines · Andrej Valek

Hello everyone,

I would like to ask you how to proceed with multiple CVEs for Google Go
component in kirkstone branch.

CVEs in current version 1.17.13:
- CVE-2022-41722
- CVE-2022-41725
- CVE-2022-41724
- CVE-2022-41723

They are fixed in 1.19.6/1.20.1 branches, but a fixing patches are available for all of them too. Unfortunately there is more then ~1000
changed LOC. So not sure if this is the right approach to apply them.
Not sure if the upgrade is acceptable.

So how to proceed with this?

I know, that they aren't a critical one, but would be nice to have them
fixed.

Regards,
Andrej

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group. View/Reply Online (#178089):
https://lists.openembedded.org/g/openembedded-core/message/178089
Mute This Topic: https://lists.openembedded.org/mt/97444547/1686489
Group Owner: [email protected] Unsubscribe:
https://lists.openembedded.org/g/openembedded-core/unsub [ [email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

#4 lore raw

Hi,

At Foundries.io we intend to update the docker version on the kirkstone branch to the latest available upstream, currently v23. Looks like the better approach can be doing that in the mixin layer, for that a new kirstone lts branch will be required.

This requires a golang update as well but the golang on master is broken for some cases when -linkshared is in use and we are still debugging this issue.
I pretend to start this backport when I can stabilize first the golang 1.20 on master.

Do you think this approach for updating docker is appropriate and acceptable?

Jose

Alexander Kanavin <[email protected]> escreveu no dia terça, 7/03/2023 à(s) 09:34:

Quoted 77 lines

If you understand the code well, and can be confident that your backports address the issue correctly and do not introduce new issues, then by all means go ahead.

My personal position should be known: I see the whole 'CVE backporting' industry as a colossal waste. We need to learn to update to supported upstream versions, and not be scared of breaking production with that.

Alex

Quoted 6 lines · Andrej

Hello Alex,

Yes, that would an option, but afaik it wasn't working quite well. So I would still prefer a straight forward solution.

Quoted 46 lines

Should I spend some time for creating such patches? Means if there will be a potential option for being accepted?

Andrej

Quoted 39 lines · Alexander Kanavin

You probably should make a kirkstone mixin layer like we did for dunfell.
https://git.yoctoproject.org/meta-lts-mixins/

Alex

Quoted 31 lines · Andrej Valek

Hello everyone,

I would like to ask you how to proceed with multiple CVEs for Google Go
component in kirkstone branch.

CVEs in current version 1.17.13:
- CVE-2022-41722
- CVE-2022-41725
- CVE-2022-41724
- CVE-2022-41723

They are fixed in 1.19.6/1.20.1 branches, but a fixing patches are available for all of them too. Unfortunately there is more then ~1000
changed LOC. So not sure if this is the right approach to apply them.
Not sure if the upgrade is acceptable.

So how to proceed with this?

I know, that they aren't a critical one, but would be nice to have them
fixed.

Regards,
Andrej

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group. View/Reply Online (#178097):
https://lists.openembedded.org/g/openembedded-core/message/178097
Mute This Topic: https://lists.openembedded.org/mt/97444547/5052612
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [ [email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Signature · 3 lines
Best regards,

José Quaresma
#5 lore raw

Hi,

Quoted 6 lines · Jose Quaresma

Hi,

At Foundries.io we intend to update the docker version on the kirkstone branch to the latest available upstream, currently v23. Looks like the better approach can be doing that in the mixin layer, for that a new kirstone lts branch will be required.

FWIW, meta-virtualization master branch works well with kirkstone and docker can be updated that way.

Many layers remove layer compatibility with older LTS releases even when technically recipes still work, at least with very few exceptions.

Quoted 5 lines

This requires a golang update as well but the golang on master is broken for some cases when -linkshared is in use and we are still debugging this issue.
I pretend to start this backport when I can stabilize first the golang 1.20 on master.

Cherry-picking go patches from poky master to kirkstone also works once the kirkstone specific security updates have been reverted.

Even if Alex disagrees, CVE patch backporting is needed when SW components break APIs and ABIs between releases. For mature and API/ABI compatible SW components the updates are trivial, but even with them there are exceptions and surprises.

I think yocto maintainers can make sensible decisions and compromises for SW components and features which get updates vs which get CVE security patches. I think many "development only" tools can also be updated quite freely
also in LTS branches. For example CVE and SPDX tooling, maybe even
bitbake itself, git, subversion etc. gcc and libc do break things on major updates, same for clang. go, rust, perl, python have possibly more stable updates so those could be aligned with all maintained branches. If no-one is posting CVE patches for complex SW components to LTS branches then I'd update them
to latest release or even master branch rather than keep the affected
CVEs open for ever. If there are not enough maintainers/resources, then even breaking the API compatibility is ok. nodejs, rust, go, webkit, tiff etc are really pain to maintain and thus updating them to latest release should be considered even if APIs change.

Cheers,

-Mikko

#6 lore raw
Quoted 6 lines · Mikko Rapeli

Hi,

Quoted 2 lines · Jose Quaresma

Hi,

Quoted 7 lines
Quoted 4 lines

At Foundries.io we intend to update the docker version on the kirkstone branch to the latest available upstream, currently v23. Looks like the better approach can be doing that in the mixin layer, for that a new kirstone lts branch will be required.

FWIW, meta-virtualization master branch works well with kirkstone and docker can be updated that way.

And I'm also happy to maintain multiple versions in the released meta-virt branches, that way we can keep compatibility, but also offer a newer version that can be selected by preferred version.

I'd rather have the main/supported recipes in the layers where most of the effort is placed, versus spraying them around in mixin layers.

But as Mikko said, also consider just trying master against the release, as chances are it is fine, and the branching is just to put a stake in the ground for when the release happened.

Bruce

Quoted 44 lines

Many layers remove layer compatibility with older LTS releases even when technically recipes still work, at least with very few exceptions.

Quoted 5 lines

This requires a golang update as well but the golang on master is broken for some cases when -linkshared is in use and we are still debugging this issue.
I pretend to start this backport when I can stabilize first the golang 1.20 on master.

Cherry-picking go patches from poky master to kirkstone also works once the kirkstone specific security updates have been reverted.

Even if Alex disagrees, CVE patch backporting is needed when SW components break APIs and ABIs between releases. For mature and API/ABI compatible SW components the updates are trivial, but even with them there are exceptions and surprises.

I think yocto maintainers can make sensible decisions and compromises for SW components and features which get updates vs which get CVE security patches. I think many "development only" tools can also be updated quite freely
also in LTS branches. For example CVE and SPDX tooling, maybe even
bitbake itself, git, subversion etc. gcc and libc do break things on major updates, same for clang. go, rust, perl, python have possibly more stable updates so those could be aligned with all maintained branches. If no-one is posting CVE patches for complex SW components to LTS branches then I'd update them
to latest release or even master branch rather than keep the affected
CVEs open for ever. If there are not enough maintainers/resources, then even breaking the API compatibility is ok. nodejs, rust, go, webkit, tiff etc are really pain to maintain and thus updating them to latest release should be considered even if APIs change.

Cheers,

-Mikko

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#178105): https://lists.openembedded.org/g/openembedded-core/message/178105
Mute This Topic: https://lists.openembedded.org/mt/97444547/1050810
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Signature · 3 lines
- Thou shalt not follow the NULL pointer, for chaos and madness await
thee at its end
- "Use the force Harry" - Gandalf, Star Trek II
#7 lore raw

Bruce Ashfield <[email protected]> escreveu no dia terça, 7/03/2023 à(s) 13:52:

Quoted 20 lines
Quoted 6 lines · Mikko Rapeli

Hi,

Quoted 2 lines · Jose Quaresma

Hi,

Quoted 7 lines
Quoted 4 lines

At Foundries.io we intend to update the docker version on the kirkstone branch to the latest available upstream, currently v23. Looks like the better approach can be doing that in the mixin layer, for that a new kirstone lts branch will be required.

FWIW, meta-virtualization master branch works well with kirkstone and docker can be updated that way.

And I'm also happy to maintain multiple versions in the released meta-virt branches, that way we can keep compatibility, but also offer a newer version that can be selected by preferred version.

Quoted 7 lines

I'd rather have the main/supported recipes in the layers where most of the effort is placed, versus spraying them around in mixin layers.

But as Mikko said, also consider just trying master against the release, as chances are it is fine, and the branching is just to put a stake in the ground for when the release happened.

One of the problems with this approach of using the master branch of meta-virt
is because some go recipes/projects require a recent version of the golang to build the latest version.
At last the golang needs to be backported on the mixin layer to circumvent these problems.

Anyway I will try to build oe-core kirkstone with meta-virt master branch to get a better overview of the build requirements.

Jose

Quoted 65 lines

Bruce

Quoted 5 lines

Many layers remove layer compatibility with older LTS releases even when technically recipes still work, at least with very few exceptions.

Quoted 1 line

This requires a golang update as well but the golang on master is

broken

Quoted 1 line
Quoted 1 line

for some cases when -linkshared is in use and we are still debugging

this

Quoted 2 lines
Quoted 2 lines

issue.
I pretend to start this backport when I can stabilize first the golang

1.20

Quoted 11 lines
Quoted 1 line

on master.

Cherry-picking go patches from poky master to kirkstone also works once the kirkstone specific security updates have been reverted.

Even if Alex disagrees, CVE patch backporting is needed when SW components break APIs and ABIs between releases. For mature and API/ABI compatible SW components the updates are trivial, but even with them there are exceptions and surprises.

I think yocto maintainers can make sensible decisions and compromises

for SW

Quoted 1 line

components and features which get updates vs which get CVE security

patches. I

Quoted 3 lines

think many "development only" tools can also be updated quite freely also in LTS branches. For example CVE and SPDX tooling, maybe even bitbake itself, git, subversion etc. gcc and libc do break things on

major updates, same for

Quoted 5 lines

clang. go, rust, perl, python have possibly more stable updates so those could be aligned with all maintained branches. If no-one is posting CVE patches for complex SW components to LTS branches then I'd update them to latest release or even master branch rather than keep the affected CVEs open for ever. If there are not enough maintainers/resources, then

even

Quoted 1 line

breaking the API compatibility is ok. nodejs, rust, go, webkit, tiff etc

are

Quoted 1 line

really pain to maintain and thus updating them to latest release should

be

Quoted 9 lines

considered even if APIs change.

Cheers,

-Mikko

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group. View/Reply Online (#178105):

https://lists.openembedded.org/g/openembedded-core/message/178105

Quoted 3 lines

Mute This Topic: https://lists.openembedded.org/mt/97444547/1050810
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [

[email protected]]

Quoted 2 lines

-=-=-=-=-=-=-=-=-=-=-=-

--
- Thou shalt not follow the NULL pointer, for chaos and madness await thee at its end
- "Use the force Harry" - Gandalf, Star Trek II

Signature · 3 lines
Best regards,

José Quaresma
#8 lore raw
Quoted 27 lines · Jose Quaresma

Bruce Ashfield <[email protected]> escreveu no dia terça, 7/03/2023 à(s) 13:52:

Quoted 23 lines · Mikko Rapeli
Quoted 6 lines

Hi,

Quoted 2 lines · Jose Quaresma

Hi,

Quoted 7 lines
Quoted 4 lines

At Foundries.io we intend to update the docker version on the kirkstone branch to the latest available upstream, currently v23. Looks like the better approach can be doing that in the mixin layer, for that a new kirstone lts branch will be required.

FWIW, meta-virtualization master branch works well with kirkstone and docker can be updated that way.

And I'm also happy to maintain multiple versions in the released meta-virt branches, that way we can keep compatibility, but also offer a newer version that can be selected by preferred version.

I'd rather have the main/supported recipes in the layers where most of the effort is placed, versus spraying them around in mixin layers.

Quoted 8 lines
Quoted 3 lines

But as Mikko said, also consider just trying master against the release, as chances are it is fine, and the branching is just to put a stake in the ground for when the release happened.

One of the problems with this approach of using the master branch of meta-virt
is because some go recipes/projects require a recent version of the golang to build the latest version. At last the golang needs to be backported on the mixin layer to circumvent these problems.

Right. You may still need some sort of mixin layer (for go, etc), I'm more saying that I would be willing to get whatever version of meta-virt tools available in the appropriate branches, since it doesn't mean a forced upgrade to all users if we just keep the multiple versions around.

Bruce

Quoted 67 lines

Anyway I will try to build oe-core kirkstone with meta-virt master branch to get a better overview of the build requirements.

Jose

Quoted 50 lines

Bruce

Quoted 36 lines

Many layers remove layer compatibility with older LTS releases even when technically recipes still work, at least with very few exceptions.

Quoted 5 lines

This requires a golang update as well but the golang on master is broken for some cases when -linkshared is in use and we are still debugging this issue.
I pretend to start this backport when I can stabilize first the golang 1.20 on master.

Cherry-picking go patches from poky master to kirkstone also works once the kirkstone specific security updates have been reverted.

Even if Alex disagrees, CVE patch backporting is needed when SW components break APIs and ABIs between releases. For mature and API/ABI compatible SW components the updates are trivial, but even with them there are exceptions and surprises.

I think yocto maintainers can make sensible decisions and compromises for SW components and features which get updates vs which get CVE security patches. I think many "development only" tools can also be updated quite freely
also in LTS branches. For example CVE and SPDX tooling, maybe even
bitbake itself, git, subversion etc. gcc and libc do break things on major updates, same for clang. go, rust, perl, python have possibly more stable updates so those could be aligned with all maintained branches. If no-one is posting CVE patches for complex SW components to LTS branches then I'd update them
to latest release or even master branch rather than keep the affected
CVEs open for ever. If there are not enough maintainers/resources, then even breaking the API compatibility is ok. nodejs, rust, go, webkit, tiff etc are really pain to maintain and thus updating them to latest release should be considered even if APIs change.

Cheers,

-Mikko

Quoted 8 lines

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#178105): https://lists.openembedded.org/g/openembedded-core/message/178105
Mute This Topic: https://lists.openembedded.org/mt/97444547/1050810
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Quoted 4 lines

--
- Thou shalt not follow the NULL pointer, for chaos and madness await thee at its end
- "Use the force Harry" - Gandalf, Star Trek II

--
Best regards,

José Quaresma

Signature · 3 lines
- Thou shalt not follow the NULL pointer, for chaos and madness await
thee at its end
- "Use the force Harry" - Gandalf, Star Trek II
#9 lore raw

Bruce Ashfield <[email protected]> escreveu no dia terça, 7/03/2023 à(s) 15:15:

Quoted 41 lines
Quoted 4 lines · Jose Quaresma

Bruce Ashfield <[email protected]> escreveu no dia terça,

7/03/2023 à(s) 13:52:

Quoted 2 lines
Quoted 2 lines

On Tue, Mar 7, 2023 at 7:15 AM Mikko Rapeli <[email protected]>

Quoted 7 lines · wrote:
Quoted 7 lines
Quoted 7 lines

Hi,

Quoted 3 lines · Jose Quaresma

Hi,

At Foundries.io we intend to update the docker version on the

kirkstone

Quoted 14 lines
Quoted 14 lines
Quoted 6 lines
Quoted 3 lines

branch to the latest available upstream, currently v23. Looks like the better approach can be doing that in the mixin layer, for that a new kirstone lts branch will be required.

FWIW, meta-virtualization master branch works well with kirkstone and docker can be updated that way.

And I'm also happy to maintain multiple versions in the released meta-virt branches, that way we can keep compatibility, but also offer a newer version that can be selected by preferred version.

I'd rather have the main/supported recipes in the layers where most of the effort is placed, versus spraying them around in mixin layers.

Quoted 6 lines
Quoted 3 lines

But as Mikko said, also consider just trying master against the release, as chances are it is fine, and the branching is just to put a stake in the ground for when the release happened.

One of the problems with this approach of using the master branch of

meta-virt

Quoted 1 line

is because some go recipes/projects require a recent version of the

golang to build the latest version.

Quoted 9 lines
Quoted 1 line

At last the golang needs to be backported on the mixin layer to

circumvent these problems.

Right. You may still need some sort of mixin layer (for go, etc), I'm more saying that I would be willing to get whatever version of meta-virt tools available in the appropriate branches, since it doesn't mean a forced upgrade to all users if we just keep the multiple versions around.

Having on meta-virt more than one version available on the stable/lts branches
looks like a great improvement. The newest version can use default_prefference -1
so we can maintain some type of abi/api stabilization enabled by default.

However it will always be difficult to decide which version of golang the tool on meta-virt needs.
And of course, the adicional maintenance overhead can increase a lot.

Jose

Quoted 93 lines

Bruce

Quoted 12 lines

Anyway I will try to build oe-core kirkstone with meta-virt master branch to get a better overview of the build requirements.

Jose

Quoted 6 lines

Bruce

Quoted 2 lines

Many layers remove layer compatibility with older LTS releases even

when

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

technically recipes still work, at least with very few exceptions.

Quoted 1 line

This requires a golang update as well but the golang on master is

broken

Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

for some cases when -linkshared is in use and we are still

debugging this

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

issue.
I pretend to start this backport when I can stabilize first the

golang 1.20

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 1 line

on master.

Cherry-picking go patches from poky master to kirkstone also works

once

Quoted 4 lines
Quoted 4 lines
Quoted 4 lines

the kirkstone specific security updates have been reverted.

Even if Alex disagrees, CVE patch backporting is needed when SW components break APIs and ABIs between releases. For mature and

API/ABI

Quoted 4 lines
Quoted 4 lines
Quoted 4 lines

compatible SW components the updates are trivial, but even with them there are exceptions and surprises.

I think yocto maintainers can make sensible decisions and compromises

for SW

Quoted 1 line
Quoted 1 line
Quoted 1 line

components and features which get updates vs which get CVE security

patches. I

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

think many "development only" tools can also be updated quite freely also in LTS branches. For example CVE and SPDX tooling, maybe even bitbake itself, git, subversion etc. gcc and libc do break things on

major updates, same for

Quoted 1 line
Quoted 1 line
Quoted 1 line

clang. go, rust, perl, python have possibly more stable updates so

those

Quoted 1 line
Quoted 1 line
Quoted 1 line

could be aligned with all maintained branches. If no-one is posting

CVE

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

patches for complex SW components to LTS branches then I'd update them to latest release or even master branch rather than keep the affected CVEs open for ever. If there are not enough maintainers/resources,

then even

Quoted 1 line
Quoted 1 line
Quoted 1 line

breaking the API compatibility is ok. nodejs, rust, go, webkit, tiff

etc are

Quoted 1 line
Quoted 1 line
Quoted 1 line

really pain to maintain and thus updating them to latest release

should be

Quoted 9 lines
Quoted 9 lines
Quoted 9 lines

considered even if APIs change.

Cheers,

-Mikko

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group. View/Reply Online (#178105):

https://lists.openembedded.org/g/openembedded-core/message/178105

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

Mute This Topic: https://lists.openembedded.org/mt/97444547/1050810
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub

[[email protected]]

Quoted 11 lines
Quoted 8 lines
Quoted 2 lines

-=-=-=-=-=-=-=-=-=-=-=-

--
- Thou shalt not follow the NULL pointer, for chaos and madness await thee at its end
- "Use the force Harry" - Gandalf, Star Trek II

Quoted 4 lines

--
Best regards,

José Quaresma

--
- Thou shalt not follow the NULL pointer, for chaos and madness await thee at its end
- "Use the force Harry" - Gandalf, Star Trek II

Signature · 3 lines
Best regards,

José Quaresma
#10 lore raw
Quoted 49 lines · Jose Quaresma

Bruce Ashfield <[email protected]> escreveu no dia terça, 7/03/2023 à(s) 15:15:

Quoted 38 lines · Jose Quaresma
Quoted 27 lines

Bruce Ashfield <[email protected]> escreveu no dia terça, 7/03/2023 à(s) 13:52:

Quoted 23 lines · Mikko Rapeli
Quoted 6 lines

Hi,

Quoted 2 lines · Jose Quaresma

Hi,

Quoted 7 lines
Quoted 4 lines

At Foundries.io we intend to update the docker version on the kirkstone branch to the latest available upstream, currently v23. Looks like the better approach can be doing that in the mixin layer, for that a new kirstone lts branch will be required.

FWIW, meta-virtualization master branch works well with kirkstone and docker can be updated that way.

And I'm also happy to maintain multiple versions in the released meta-virt branches, that way we can keep compatibility, but also offer a newer version that can be selected by preferred version.

I'd rather have the main/supported recipes in the layers where most of the effort is placed, versus spraying them around in mixin layers.

Quoted 8 lines
Quoted 3 lines

But as Mikko said, also consider just trying master against the release, as chances are it is fine, and the branching is just to put a stake in the ground for when the release happened.

One of the problems with this approach of using the master branch of meta-virt
is because some go recipes/projects require a recent version of the golang to build the latest version. At last the golang needs to be backported on the mixin layer to circumvent these problems.

Quoted 5 lines

Right. You may still need some sort of mixin layer (for go, etc), I'm more saying that I would be willing to get whatever version of meta-virt tools available in the appropriate branches, since it doesn't mean a forced upgrade to all users if we just keep the multiple versions around.

Quoted 6 lines

Having on meta-virt more than one version available on the stable/lts branches looks like a great improvement. The newest version can use default_prefference -1 so we can maintain some type of abi/api stabilization enabled by default.

However it will always be difficult to decide which version of golang the tool on meta-virt needs. And of course, the adicional maintenance overhead can increase a lot.

Having battled the go applications in meta-virt for a long time now, it really isn't all that often we have issues with golang versions, in the sense of needing matched versions. It is normally new features that are required, versus a new golang dropping features. So a newer go has almost always worked for all the versions we need.

The extra maintenance isn't significant, in particular when balanced around trying to keep a centre of gravity around the recipes.

Bruce

Quoted 87 lines

Jose

Quoted 69 lines

Bruce

Quoted 66 lines

Anyway I will try to build oe-core kirkstone with meta-virt master branch to get a better overview of the build requirements.

Jose

Quoted 50 lines

Bruce

Quoted 36 lines

Many layers remove layer compatibility with older LTS releases even when technically recipes still work, at least with very few exceptions.

Quoted 5 lines

This requires a golang update as well but the golang on master is broken for some cases when -linkshared is in use and we are still debugging this issue.
I pretend to start this backport when I can stabilize first the golang 1.20 on master.

Cherry-picking go patches from poky master to kirkstone also works once the kirkstone specific security updates have been reverted.

Even if Alex disagrees, CVE patch backporting is needed when SW components break APIs and ABIs between releases. For mature and API/ABI compatible SW components the updates are trivial, but even with them there are exceptions and surprises.

I think yocto maintainers can make sensible decisions and compromises for SW components and features which get updates vs which get CVE security patches. I think many "development only" tools can also be updated quite freely
also in LTS branches. For example CVE and SPDX tooling, maybe even
bitbake itself, git, subversion etc. gcc and libc do break things on major updates, same for clang. go, rust, perl, python have possibly more stable updates so those could be aligned with all maintained branches. If no-one is posting CVE patches for complex SW components to LTS branches then I'd update them
to latest release or even master branch rather than keep the affected
CVEs open for ever. If there are not enough maintainers/resources, then even breaking the API compatibility is ok. nodejs, rust, go, webkit, tiff etc are really pain to maintain and thus updating them to latest release should be considered even if APIs change.

Cheers,

-Mikko

Quoted 8 lines

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#178105): https://lists.openembedded.org/g/openembedded-core/message/178105
Mute This Topic: https://lists.openembedded.org/mt/97444547/1050810
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Quoted 4 lines

--
- Thou shalt not follow the NULL pointer, for chaos and madness await thee at its end
- "Use the force Harry" - Gandalf, Star Trek II

--
Best regards,

Quoted 8 lines
Quoted 1 line

José Quaresma

--
- Thou shalt not follow the NULL pointer, for chaos and madness await thee at its end
- "Use the force Harry" - Gandalf, Star Trek II

--
Best regards,

José Quaresma

Signature · 3 lines
- Thou shalt not follow the NULL pointer, for chaos and madness await
thee at its end
- "Use the force Harry" - Gandalf, Star Trek II
#11 lore raw

Hi,

From my perspective, the challenge (and also the original question of this thread) is the short lifecycle of the newer language stacks, namely Golang and Rust. Security patches are not provided by the go community for the entire LTS period of Yocto. Golang practices are partially at odds with the idea of a LTS distro. It's not only about Docker or other components, it's also about the go tool-chain itself which provides a ~50MB library (with vulnarabilities) which gets linked (usually statically and stripped) by all go applications.

When we were still on the Dunfell branch, we had tried to solve the issue with the mixin repo from
https://git.yoctoproject.org/meta-lts-mixins/refs/. But it didn't really work well. The mixin layer brings the latest go stack from poky master back to poky dunfell. Since there are no go recipes in poky so far, this works relatively easily. But adding this layer more or less breaks all recipes from meta-virtualization. A mixin layer which updates the main go tool-chain does not look like a generic and easy maintainable approach.

On the other hand, Golang and Rust are much more self-contained than other language stacks, which in turn makes it relatively easy to install more than one Golang version in parallel on a host and to use one or the other Golang version depending on the software to be compiled. The development approaches of most go-based software components are also based on this assumption. Docker 21, for example, had required go 1.17 and then switched to go 1.18. At least my conclusion was that the switch to golang 1.18 must be reverted to allow compiling the latest Docker 21 with golang 1.17 from poky. See
https://git.yoctoproject.org/meta-virtualization/commit/?h=kirkstone&id=927537108bcf2b98859512ce3eae59a73439994d
. This means that as long as there is only one go version available with poky, updating or maintaining multiple version of go based recipes is only possible for some recipes and comes with additional effort per recipe.

My conclusion so far is that the most generic and probably also the most straight forward approach would be to support more than one go (and probably also rust) version for the LTS branches. If recipes could choose the go-native version for example meta-virtualization could offer e.g. Docker 21 with DEPENDS += "go-native" and also Docker 23 with DEPENDS += "go-1-19-native". Please note that I'm thinking about different go-native recipes with the major version of go in the recipe name not about different go-native recipe versions. This would allow to make them concurently available. Additional go versions could be provided by poky or by a mixin layer.

This approach would make it possible to keep go recipes more up-to-date and thus also fix security vulnerabilities without having to backport many patches to many go-dependent recipes. This would probably be ideal for meta-virtualization (especially if multiple versions should be supported for certain recipes) as well as for many practical scenarios.

But at least in the first place it would not help to improve the situation of open CVEs in upstream unmaintained go tool-chains which are still used by poky LTS. Backported patches would still be needed to keep poky itself secure and get nice CVE metrics for poky world builds until older tool-chains would be replaced or at least somehow deprecated on the LTS branches.

Regards Adrian