Re: [OE-Core][PATCH v14 03/11] sstate.bbclass: override tar with tar-native

["Piotr Łobacz" <p.lobacz@welotec.com>]

Dnia poniedziałek, 16 października 2023 07:45:41 CET Piotr Łobacz via 
lists.openembedded.org pisze:
> Dnia niedziela, 15 października 2023 20:26:49 CEST Richard Purdie pisze:
> > On Sun, 2023-10-15 at 16:57 +0200, Piotr Łobacz wrote:
> > > Due to the bugs
> > > https://eur04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fsavann
> > > a%2F&data=05%7C01%7Cp.lobacz%40welotec.com%7Cbac3f04a07664b3a912a08dbce0
> > > b282c%7C25111a7f1d5a4c51a4ca7f8e44011b39%7C0%7C0%7C638330319545808351%7C
> > > Unknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haW
> > > wiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=875wwpxmvY10%2BUdvEbJFTJwc1zW%2Bq
> > > 0YOn8g9wgVvdUA%3D&reserved=0
> > > h.gnu.org%2Fbugs%2F%3F61934&data=05%7C01%7Cp.lobacz%40welotec.com%7C208
> > > 43c
> > > d409ff471ad53b08dbcdac4d55%7C25111a7f1d5a4c51a4ca7f8e44011b39%7C0%7C0%7
> > > C63
> > > 8329912144020487%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2
> > > luM
> > > zIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=U8%2Fmm02MR5ps5
> > > btR JnE1EnySNk4NJV498kRX66jjeeQ%3D&reserved=0 and
> > > https://eur04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fsavann
> > > a%2F&data=05%7C01%7Cp.lobacz%40welotec.com%7Cbac3f04a07664b3a912a08dbce0
> > > b282c%7C25111a7f1d5a4c51a4ca7f8e44011b39%7C0%7C0%7C638330319545808351%7C
> > > Unknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haW
> > > wiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=875wwpxmvY10%2BUdvEbJFTJwc1zW%2Bq
> > > 0YOn8g9wgVvdUA%3D&reserved=0
> > > h.gnu.org%2Fbugs%2F%3F59184&data=05%7C01%7Cp.lobacz%40welotec.com%7C208
> > > 43c
> > > d409ff471ad53b08dbcdac4d55%7C25111a7f1d5a4c51a4ca7f8e44011b39%7C0%7C0%7
> > > C63
> > > 8329912144020487%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2
> > > luM
> > > zIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=J3cwWWAQ6BHP1aG
> > > VeP 1yRIGnGyG1fqVvPaVJY6kjKWc%3D&reserved=0, which are fixed in tar
> > > version 1.35, we need to force yocto to use it instead
> > > of the one provided by the host machine.
> > > 
> > > sstate_create_package and sstate_unpack_package functions,
> > > will use the tar provided by tar recipe, which is already
> > > in proper version.
> > > 
> > > Signed-off-by: Piotr Łobacz <p.lobacz@welotec.com>
> > > ---
> > > 
> > >  meta/classes-global/sstate.bbclass | 6 +++++-
> > >  1 file changed, 5 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/meta/classes-global/sstate.bbclass
> > > b/meta/classes-global/sstate.bbclass index 2676f18e0a..5ae6f515f2 100644
> > > --- a/meta/classes-global/sstate.bbclass
> > > +++ b/meta/classes-global/sstate.bbclass
> > > @@ -152,7 +152,9 @@ python () {
> > > 
> > >      else:
> > >          d.setVar('SSTATE_MANMACH', d.expand("${PACKAGE_ARCH}"))
> > > 
> > > -    if bb.data.inherits_class('native', d) or
> > > bb.data.inherits_class('crosssdk', d) or bb.data.inherits_class('cross',
> > > d): +    is_native = bb.data.inherits_class('native', d) or
> > > bb.data.inherits_class('crosssdk', d) or bb.data.inherits_class('cross',
> > > d) +
> > > 
> > > +    if is_native:
> > >          d.setVar('SSTATE_EXTRAPATH', "${NATIVELSBSTRING}/")
> > >          d.setVar('BB_HASHFILENAME', "True ${SSTATE_PKGSPEC}
> > >          ${SSTATE_SWSPEC}")
> > >          d.setVar('SSTATE_EXTRAPATHWILDCARD', "${NATIVELSBSTRING}/")
> > > 
> > > @@ -164,6 +166,8 @@ python () {
> > > 
> > >          d.appendVarFlag(task, 'postfuncs', " sstate_task_postfunc")
> > >          d.setVarFlag(task, 'network', '1')
> > >          d.setVarFlag(task + "_setscene", 'network', '1')
> > > 
> > > +        if not is_native:
> > > +            d.appendVarFlag(task, 'deptask', '
> > > tar-replacement-native:do_populate_sysroot')>
> > > 
> > >  }
> > 
> > >  def sstate_init(task, d):
> > I don't think this dependency is enough to ensure that the setscene
> > task for one of these tasks has tar-native present when the sstate
> > package is extracted from sstate?
> > 
> > Cheers,
> > 
> > Richard
> 
> Hi Richard, proposals are welcome :-)
> 
> BR
> Piotr
BTW. Richard was it tested already by you?

BR
Piotr