[warrior][PATCH] bitbake.conf: add git-lfs to HOSTTOOLS_NONFATAL

[warrior][PATCH] bitbake.conf: add git-lfs to HOSTTOOLS_NONFATAL

[Naveen Saini]

This provides git large file storage (lfs) extension.

Include git-lfs conditionally. If git-lfs is present on host and repo
has lfs pointers, then git-lfs will be used. If git-lfs is not present
on host, it will be ignored.

[YOCTO #13198]

(From OE-Core rev: 2968ad8514721ec06e67aaf3fd5ec7b247b3431d)

Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
---
 meta/conf/bitbake.conf | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/meta/conf/bitbake.conf b/meta/conf/bitbake.conf
index c996301a8b..c5313ccd19 100644
--- a/meta/conf/bitbake.conf
+++ b/meta/conf/bitbake.conf
@@ -514,6 +514,9 @@ HOSTTOOLS_NONFATAL += "bzr"
 # Used by ssh fetcher
 HOSTTOOLS_NONFATAL += "scp"
 
+# Link to git-lfs if present
+HOSTTOOLS_NONFATAL += "git-lfs"
+
 CCACHE ??= ""
 
 TOOLCHAIN_OPTIONS = " --sysroot=${STAGING_DIR_TARGET}"
-- 
2.17.1

Re: [warrior][PATCH] bitbake.conf: add git-lfs to HOSTTOOLS_NONFATAL

[Saini, Naveen Kumar]

The corresponding first patch of this patch series already merged in 
Warrior

https://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/bitbake/lib?h=warrior&id=6e3a4d7926296380da23536c29af35d5702e02fb

So this patch should also go in warrior.


On 8/15/19 2:53 PM, Naveen Saini wrote:
> This provides git large file storage (lfs) extension.
> 
> Include git-lfs conditionally. If git-lfs is present on host and repo
> has lfs pointers, then git-lfs will be used. If git-lfs is not present
> on host, it will be ignored.
> 
> [YOCTO #13198]
> 
> (From OE-Core rev: 2968ad8514721ec06e67aaf3fd5ec7b247b3431d)
> 
> Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
> ---
>   meta/conf/bitbake.conf | 3 +++
>   1 file changed, 3 insertions(+)
> 
> diff --git a/meta/conf/bitbake.conf b/meta/conf/bitbake.conf
> index c996301a8b..c5313ccd19 100644
> --- a/meta/conf/bitbake.conf
> +++ b/meta/conf/bitbake.conf
> @@ -514,6 +514,9 @@ HOSTTOOLS_NONFATAL += "bzr"
>   # Used by ssh fetcher
>   HOSTTOOLS_NONFATAL += "scp"
>   
> +# Link to git-lfs if present
> +HOSTTOOLS_NONFATAL += "git-lfs"
> +
>   CCACHE ??= ""
>   
>   TOOLCHAIN_OPTIONS = " --sysroot=${STAGING_DIR_TARGET}"
> 

Re: [warrior][PATCH] bitbake.conf: add git-lfs to HOSTTOOLS_NONFATAL

[Martin Jansa]

[-- Attachment #1: Type: text/plain, Size: 2039 bytes --]

NAK

Yes, the first part was merged in warrior and is correct.

But this second part isn't good (you don't want git-lfs to sometimes work
and sometimes fail) and that's why it was rejected for master and
_shouldn't_ be merged to warrior. If you have recipes which need git-lfs,
then add it to normal HOSTTOOLS in your builds to make sure it's always
present when needed.

On Thu, Aug 15, 2019 at 9:01 AM Saini, Naveen Kumar <
naveen.kumar.saini@intel.com> wrote:

> The corresponding first patch of this patch series already merged in
> Warrior
>
>
> https://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/bitbake/lib?h=warrior&id=6e3a4d7926296380da23536c29af35d5702e02fb
>
> So this patch should also go in warrior.
>
>
> On 8/15/19 2:53 PM, Naveen Saini wrote:
> > This provides git large file storage (lfs) extension.
> >
> > Include git-lfs conditionally. If git-lfs is present on host and repo
> > has lfs pointers, then git-lfs will be used. If git-lfs is not present
> > on host, it will be ignored.
> >
> > [YOCTO #13198]
> >
> > (From OE-Core rev: 2968ad8514721ec06e67aaf3fd5ec7b247b3431d)
> >
> > Signed-off-by: Naveen Saini <naveen.kumar.saini@intel.com>
> > Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
> > ---
> >   meta/conf/bitbake.conf | 3 +++
> >   1 file changed, 3 insertions(+)
> >
> > diff --git a/meta/conf/bitbake.conf b/meta/conf/bitbake.conf
> > index c996301a8b..c5313ccd19 100644
> > --- a/meta/conf/bitbake.conf
> > +++ b/meta/conf/bitbake.conf
> > @@ -514,6 +514,9 @@ HOSTTOOLS_NONFATAL += "bzr"
> >   # Used by ssh fetcher
> >   HOSTTOOLS_NONFATAL += "scp"
> >
> > +# Link to git-lfs if present
> > +HOSTTOOLS_NONFATAL += "git-lfs"
> > +
> >   CCACHE ??= ""
> >
> >   TOOLCHAIN_OPTIONS = " --sysroot=${STAGING_DIR_TARGET}"
> >
> --
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core
>

[-- Attachment #2: Type: text/html, Size: 3158 bytes --]

Re: [warrior][PATCH] bitbake.conf: add git-lfs to HOSTTOOLS_NONFATAL

[Saini, Naveen Kumar]

On 8/15/19 3:23 PM, Martin Jansa wrote:
> NAK
> 
> Yes, the first part was merged in warrior and is correct.
> 
> But this second part isn't good (you don't want git-lfs to sometimes 
> work and sometimes fail) and that's why it was rejected for master and 
> _shouldn't_ be merged to warrior. If you have recipes which need 
> git-lfs, then add it to normal HOSTTOOLS in your builds to make sure 
> it's always present when needed.

Noted. Thanks.

Re: [warrior][PATCH] bitbake.conf: add git-lfs to HOSTTOOLS_NONFATAL

[Richard Purdie]

On Thu, 2019-08-15 at 09:23 +0200, Martin Jansa wrote:
> NAK
> 
> Yes, the first part was merged in warrior and is correct.
> 
> But this second part isn't good (you don't want git-lfs to sometimes
> work and sometimes fail) and that's why it was rejected for master
> and _shouldn't_ be merged to warrior. If you have recipes which need
> git-lfs, then add it to normal HOSTTOOLS in your builds to make sure
> it's always present when needed.

I don't like the patch but it did merge to master.

Building a git-lfs-native is a nightmare due to all its dependencies (I
think its go based?) and there wasn't really any other way to sort it.

Having recipe specific hosttools would be ideal but we don't have that.

Cheers,

Richard

Re: [warrior][PATCH] bitbake.conf: add git-lfs to HOSTTOOLS_NONFATAL

[Andre McCurdy]

On Thu, Aug 15, 2019 at 2:09 PM Richard Purdie
<richard.purdie@linuxfoundation.org> wrote:
> On Thu, 2019-08-15 at 09:23 +0200, Martin Jansa wrote:
> > NAK
> >
> > Yes, the first part was merged in warrior and is correct.
> >
> > But this second part isn't good (you don't want git-lfs to sometimes
> > work and sometimes fail) and that's why it was rejected for master
> > and _shouldn't_ be merged to warrior. If you have recipes which need
> > git-lfs, then add it to normal HOSTTOOLS in your builds to make sure
> > it's always present when needed.
>
> I don't like the patch but it did merge to master.
>
> Building a git-lfs-native is a nightmare due to all its dependencies (I
> think its go based?) and there wasn't really any other way to sort it.

Upstream provides generic binaries for at least x86, x86-64 and ARM64:

  https://github.com/git-lfs/git-lfs/releases
  https://github.com/git-lfs/git-lfs/releases/download/v2.8.0/git-lfs-linux-386-v2.8.0.tar.gz
  https://github.com/git-lfs/git-lfs/releases/download/v2.8.0/git-lfs-linux-amd64-v2.8.0.tar.gz
  https://github.com/git-lfs/git-lfs/releases/download/v2.8.0/git-lfs-linux-arm64-v2.8.0.tar.gz

Would just downloading and installing the official upstream binaries be so bad?

> Having recipe specific hosttools would be ideal but we don't have that.
>
> Cheers,
>
> Richard
>
> --
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core

Re: [warrior][PATCH] bitbake.conf: add git-lfs to HOSTTOOLS_NONFATAL

[akuster808]

On 8/15/19 2:31 PM, Andre McCurdy wrote:
> On Thu, Aug 15, 2019 at 2:09 PM Richard Purdie
> <richard.purdie@linuxfoundation.org> wrote:
>> On Thu, 2019-08-15 at 09:23 +0200, Martin Jansa wrote:
>>> NAK
>>>
>>> Yes, the first part was merged in warrior and is correct.
>>>
>>> But this second part isn't good (you don't want git-lfs to sometimes
>>> work and sometimes fail) and that's why it was rejected for master
>>> and _shouldn't_ be merged to warrior. If you have recipes which need
>>> git-lfs, then add it to normal HOSTTOOLS in your builds to make sure
>>> it's always present when needed.
>> I don't like the patch but it did merge to master.
>>
>> Building a git-lfs-native is a nightmare due to all its dependencies (I
>> think its go based?) and there wasn't really any other way to sort it.
> Upstream provides generic binaries for at least x86, x86-64 and ARM64:
>
>   https://github.com/git-lfs/git-lfs/releases
>   https://github.com/git-lfs/git-lfs/releases/download/v2.8.0/git-lfs-linux-386-v2.8.0.tar.gz
>   https://github.com/git-lfs/git-lfs/releases/download/v2.8.0/git-lfs-linux-amd64-v2.8.0.tar.gz
>   https://github.com/git-lfs/git-lfs/releases/download/v2.8.0/git-lfs-linux-arm64-v2.8.0.tar.gz
>
> Would just downloading and installing the official upstream binaries be so bad?

making a big change like this seems to be a no-go for a stable branch.

- armin
>
>> Having recipe specific hosttools would be ideal but we don't have that.
>>
>> Cheers,
>>
>> Richard
>>
>> --
>> _______________________________________________
>> Openembedded-core mailing list
>> Openembedded-core@lists.openembedded.org
>> http://lists.openembedded.org/mailman/listinfo/openembedded-core

Re: [warrior][PATCH] bitbake.conf: add git-lfs to HOSTTOOLS_NONFATAL

[Andre McCurdy]

On Thu, Aug 15, 2019 at 8:45 PM akuster808 <akuster808@gmail.com> wrote:
> On 8/15/19 2:31 PM, Andre McCurdy wrote:
> > On Thu, Aug 15, 2019 at 2:09 PM Richard Purdie
> > <richard.purdie@linuxfoundation.org> wrote:
> >> On Thu, 2019-08-15 at 09:23 +0200, Martin Jansa wrote:
> >>> NAK
> >>>
> >>> Yes, the first part was merged in warrior and is correct.
> >>>
> >>> But this second part isn't good (you don't want git-lfs to sometimes
> >>> work and sometimes fail) and that's why it was rejected for master
> >>> and _shouldn't_ be merged to warrior. If you have recipes which need
> >>> git-lfs, then add it to normal HOSTTOOLS in your builds to make sure
> >>> it's always present when needed.
> >> I don't like the patch but it did merge to master.
> >>
> >> Building a git-lfs-native is a nightmare due to all its dependencies (I
> >> think its go based?) and there wasn't really any other way to sort it.
> > Upstream provides generic binaries for at least x86, x86-64 and ARM64:
> >
> >   https://github.com/git-lfs/git-lfs/releases
> >   https://github.com/git-lfs/git-lfs/releases/download/v2.8.0/git-lfs-linux-386-v2.8.0.tar.gz
> >   https://github.com/git-lfs/git-lfs/releases/download/v2.8.0/git-lfs-linux-amd64-v2.8.0.tar.gz
> >   https://github.com/git-lfs/git-lfs/releases/download/v2.8.0/git-lfs-linux-arm64-v2.8.0.tar.gz
> >
> > Would just downloading and installing the official upstream binaries be so bad?
>
> making a big change like this seems to be a no-go for a stable branch.

Right. This probably isn't suitable for the stable branch. It would
obviously be implemented first in master anyway.

Point was just that if we made use of the upstream binaries (which are
completely distro independent as far as I can tell?) then a
git-lfs-native recipe would transform from "a nightmare" to almost
trivial.