diff mbox series

[meta-networking,scarthgap] mbedtls: fix git fetcher conflict by using SRC_URI.

Message ID 20250815041114.481650-1-karn.jye.lau@intel.com
State New
Headers show
Series [meta-networking,scarthgap] mbedtls: fix git fetcher conflict by using SRC_URI. | expand

Commit Message

Lau, Karn Jye Aug. 15, 2025, 4:11 a.m. UTC
From: kjlau0112 <karn.jye.lau@intel.com>

if both SRC_URI & SRCREV specify particular source revision
will lead to "Fix Fetcher failure: Conflicting revisions error"

Fix by removing SRCREV and specify expected source revision in SRC_URI
by adding revision parmater.

Signed-off-by: kjlau0112 <karn.jye.lau@intel.com>
---
 meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

Comments

Peter Marko Aug. 15, 2025, 5:52 a.m. UTC | #1
Both tag and SRVREV works with latest oe-core/poky master.
And is actually encouraged to be used so that hash is verified against the tag.

I wonder what could be wrong since this is a second patch for this recipe complaining about errors.

Peter

> -----Original Message-----
> From: openembedded-devel@lists.openembedded.org <openembedded-
> devel@lists.openembedded.org> On Behalf Of Lau, Karn Jye via
> lists.openembedded.org
> Sent: Friday, August 15, 2025 6:11
> To: openembedded-devel@lists.openembedded.org
> Subject: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git fetcher conflict
> by using SRC_URI.
> 
> From: kjlau0112 <karn.jye.lau@intel.com>
> 
> if both SRC_URI & SRCREV specify particular source revision
> will lead to "Fix Fetcher failure: Conflicting revisions error"
> 
> Fix by removing SRCREV and specify expected source revision in SRC_URI
> by adding revision parmater.
> 
> Signed-off-by: kjlau0112 <karn.jye.lau@intel.com>
> ---
>  meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb | 4 +---
>  1 file changed, 1 insertion(+), 3 deletions(-)
> 
> diff --git a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb b/meta-
> networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> index 88b50a406..2591e2cde 100644
> --- a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> +++ b/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> @@ -23,12 +23,10 @@ LIC_FILES_CHKSUM =
> "file://LICENSE;md5=379d5819937a6c2f1ef1630d341e026d"
>  SECTION = "libs"
> 
>  S = "${WORKDIR}/git"
> -SRC_URI = "gitsm://github.com/Mbed-
> TLS/mbedtls.git;protocol=https;branch=mbedtls-3.6;tag=v${PV} \
> +SRC_URI = "gitsm://github.com/Mbed-
> TLS/mbedtls.git;protocol=https;branch=mbedtls-
> 3.6;rev=c765c831e5c2a0971410692f92f7a81d6ec65ec2\
>             file://run-ptest \
>             "
> 
> -SRCREV = "c765c831e5c2a0971410692f92f7a81d6ec65ec2"
> -
>  UPSTREAM_CHECK_GITTAGREGEX = "v(?P<pver>\d+(\.\d+)+)"
> 
>  inherit cmake update-alternatives ptest
> --
> 2.34.1
Lau, Karn Jye Aug. 15, 2025, 6:09 a.m. UTC | #2
Hi Peter, it may working fine in master branch
but it not working at all at LTS branch for example yocto 5.0 scargapth, I observing our daily build hit into this issue.
Thus submitting this patch

KJ
-----Original Message-----
From: Marko, Peter <Peter.Marko@siemens.com> 
Sent: Friday, August 15, 2025 1:53 PM
To: Lau, Karn Jye <karn.jye.lau@intel.com>; openembedded-devel@lists.openembedded.org
Subject: RE: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git fetcher conflict by using SRC_URI.

Both tag and SRVREV works with latest oe-core/poky master.
And is actually encouraged to be used so that hash is verified against the tag.

I wonder what could be wrong since this is a second patch for this recipe complaining about errors.

Peter

> -----Original Message-----
> From: openembedded-devel@lists.openembedded.org <openembedded- 
> devel@lists.openembedded.org> On Behalf Of Lau, Karn Jye via 
> lists.openembedded.org
> Sent: Friday, August 15, 2025 6:11
> To: openembedded-devel@lists.openembedded.org
> Subject: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git 
> fetcher conflict by using SRC_URI.
> 
> From: kjlau0112 <karn.jye.lau@intel.com>
> 
> if both SRC_URI & SRCREV specify particular source revision will lead 
> to "Fix Fetcher failure: Conflicting revisions error"
> 
> Fix by removing SRCREV and specify expected source revision in SRC_URI 
> by adding revision parmater.
> 
> Signed-off-by: kjlau0112 <karn.jye.lau@intel.com>
> ---
>  meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb | 4 
> +---
>  1 file changed, 1 insertion(+), 3 deletions(-)
> 
> diff --git 
> a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb 
> b/meta- networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> index 88b50a406..2591e2cde 100644
> --- a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> +++ b/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> @@ -23,12 +23,10 @@ LIC_FILES_CHKSUM = 
> "file://LICENSE;md5=379d5819937a6c2f1ef1630d341e026d"
>  SECTION = "libs"
> 
>  S = "${WORKDIR}/git"
> -SRC_URI = "gitsm://github.com/Mbed-
> TLS/mbedtls.git;protocol=https;branch=mbedtls-3.6;tag=v${PV} \
> +SRC_URI = "gitsm://github.com/Mbed-
> TLS/mbedtls.git;protocol=https;branch=mbedtls-
> 3.6;rev=c765c831e5c2a0971410692f92f7a81d6ec65ec2\
>             file://run-ptest \
>             "
> 
> -SRCREV = "c765c831e5c2a0971410692f92f7a81d6ec65ec2"
> -
>  UPSTREAM_CHECK_GITTAGREGEX = "v(?P<pver>\d+(\.\d+)+)"
> 
>  inherit cmake update-alternatives ptest
> --
> 2.34.1
Peter Marko Aug. 15, 2025, 6:12 a.m. UTC | #3
Oops, sorry for missing the scarthgap tag.
This is good patch, I wonder how this got through the testing.

Unfortunately, meta-openembedded LTS branches get updates only once a month so this will be broken trough then next release.

Peter

> -----Original Message-----
> From: Lau, Karn Jye <karn.jye.lau@intel.com>
> Sent: Friday, August 15, 2025 8:09
> To: Marko, Peter (FT D EU SK BFS1) <Peter.Marko@siemens.com>;
> openembedded-devel@lists.openembedded.org
> Subject: RE: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git fetcher
> conflict by using SRC_URI.
> 
> Hi Peter, it may working fine in master branch
> but it not working at all at LTS branch for example yocto 5.0 scargapth, I observing
> our daily build hit into this issue.
> Thus submitting this patch
> 
> KJ
> -----Original Message-----
> From: Marko, Peter <Peter.Marko@siemens.com>
> Sent: Friday, August 15, 2025 1:53 PM
> To: Lau, Karn Jye <karn.jye.lau@intel.com>; openembedded-
> devel@lists.openembedded.org
> Subject: RE: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git fetcher
> conflict by using SRC_URI.
> 
> Both tag and SRVREV works with latest oe-core/poky master.
> And is actually encouraged to be used so that hash is verified against the tag.
> 
> I wonder what could be wrong since this is a second patch for this recipe
> complaining about errors.
> 
> Peter
> 
> > -----Original Message-----
> > From: openembedded-devel@lists.openembedded.org <openembedded-
> > devel@lists.openembedded.org> On Behalf Of Lau, Karn Jye via
> > lists.openembedded.org
> > Sent: Friday, August 15, 2025 6:11
> > To: openembedded-devel@lists.openembedded.org
> > Subject: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git
> > fetcher conflict by using SRC_URI.
> >
> > From: kjlau0112 <karn.jye.lau@intel.com>
> >
> > if both SRC_URI & SRCREV specify particular source revision will lead
> > to "Fix Fetcher failure: Conflicting revisions error"
> >
> > Fix by removing SRCREV and specify expected source revision in SRC_URI
> > by adding revision parmater.
> >
> > Signed-off-by: kjlau0112 <karn.jye.lau@intel.com>
> > ---
> >  meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb | 4
> > +---
> >  1 file changed, 1 insertion(+), 3 deletions(-)
> >
> > diff --git
> > a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> > b/meta- networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> > index 88b50a406..2591e2cde 100644
> > --- a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> > +++ b/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> > @@ -23,12 +23,10 @@ LIC_FILES_CHKSUM =
> > "file://LICENSE;md5=379d5819937a6c2f1ef1630d341e026d"
> >  SECTION = "libs"
> >
> >  S = "${WORKDIR}/git"
> > -SRC_URI = "gitsm://github.com/Mbed-
> > TLS/mbedtls.git;protocol=https;branch=mbedtls-3.6;tag=v${PV} \
> > +SRC_URI = "gitsm://github.com/Mbed-
> > TLS/mbedtls.git;protocol=https;branch=mbedtls-
> > 3.6;rev=c765c831e5c2a0971410692f92f7a81d6ec65ec2\
> >             file://run-ptest \
> >             "
> >
> > -SRCREV = "c765c831e5c2a0971410692f92f7a81d6ec65ec2"
> > -
> >  UPSTREAM_CHECK_GITTAGREGEX = "v(?P<pver>\d+(\.\d+)+)"
> >
> >  inherit cmake update-alternatives ptest
> > --
> > 2.34.1
Peter Marko Aug. 15, 2025, 8:09 a.m. UTC | #4
> -----Original Message-----
> From: Marko, Peter (FT D EU SK BFS1)
> Sent: Friday, August 15, 2025 8:12
> To: Lau, Karn Jye <karn.jye.lau@intel.com>; openembedded-
> devel@lists.openembedded.org
> Subject: RE: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git fetcher
> conflict by using SRC_URI.
> 
> Oops, sorry for missing the scarthgap tag.
> This is good patch, I wonder how this got through the testing.
> 
> Unfortunately, meta-openembedded LTS branches get updates only once a month
> so this will be broken trough then next release.
> 
> Peter
> 
> > -----Original Message-----
> > From: Lau, Karn Jye <karn.jye.lau@intel.com>
> > Sent: Friday, August 15, 2025 8:09
> > To: Marko, Peter (FT D EU SK BFS1) <Peter.Marko@siemens.com>;
> > openembedded-devel@lists.openembedded.org
> > Subject: RE: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git fetcher
> > conflict by using SRC_URI.
> >
> > Hi Peter, it may working fine in master branch
> > but it not working at all at LTS branch for example yocto 5.0 scargapth, I
> observing
> > our daily build hit into this issue.
> > Thus submitting this patch
> >
> > KJ
> > -----Original Message-----
> > From: Marko, Peter <Peter.Marko@siemens.com>
> > Sent: Friday, August 15, 2025 1:53 PM
> > To: Lau, Karn Jye <karn.jye.lau@intel.com>; openembedded-
> > devel@lists.openembedded.org
> > Subject: RE: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git fetcher
> > conflict by using SRC_URI.
> >
> > Both tag and SRVREV works with latest oe-core/poky master.
> > And is actually encouraged to be used so that hash is verified against the tag.
> >
> > I wonder what could be wrong since this is a second patch for this recipe
> > complaining about errors.
> >
> > Peter
> >
> > > -----Original Message-----
> > > From: openembedded-devel@lists.openembedded.org <openembedded-
> > > devel@lists.openembedded.org> On Behalf Of Lau, Karn Jye via
> > > lists.openembedded.org
> > > Sent: Friday, August 15, 2025 6:11
> > > To: openembedded-devel@lists.openembedded.org
> > > Subject: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git
> > > fetcher conflict by using SRC_URI.
> > >
> > > From: kjlau0112 <karn.jye.lau@intel.com>
> > >
> > > if both SRC_URI & SRCREV specify particular source revision will lead
> > > to "Fix Fetcher failure: Conflicting revisions error"
> > >
> > > Fix by removing SRCREV and specify expected source revision in SRC_URI
> > > by adding revision parmater.
> > >
> > > Signed-off-by: kjlau0112 <karn.jye.lau@intel.com>
> > > ---
> > >  meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb | 4
> > > +---
> > >  1 file changed, 1 insertion(+), 3 deletions(-)
> > >
> > > diff --git
> > > a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> > > b/meta- networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> > > index 88b50a406..2591e2cde 100644
> > > --- a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> > > +++ b/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> > > @@ -23,12 +23,10 @@ LIC_FILES_CHKSUM =
> > > "file://LICENSE;md5=379d5819937a6c2f1ef1630d341e026d"
> > >  SECTION = "libs"
> > >
> > >  S = "${WORKDIR}/git"
> > > -SRC_URI = "gitsm://github.com/Mbed-
> > > TLS/mbedtls.git;protocol=https;branch=mbedtls-3.6;tag=v${PV} \
> > > +SRC_URI = "gitsm://github.com/Mbed-
> > > TLS/mbedtls.git;protocol=https;branch=mbedtls-
> > > 3.6;rev=c765c831e5c2a0971410692f92f7a81d6ec65ec2\
> > >             file://run-ptest \
> > >             "
> > >
> > > -SRCREV = "c765c831e5c2a0971410692f92f7a81d6ec65ec2"

Could you please keep the SRCREV and just remove the ";tag=v${PV}"?
Adding rev to SRC_URI is not according to current standards.

Peter

> > > -
> > >  UPSTREAM_CHECK_GITTAGREGEX = "v(?P<pver>\d+(\.\d+)+)"
> > >
> > >  inherit cmake update-alternatives ptest
> > > --
> > > 2.34.1
Alexander Kanavin Aug. 15, 2025, 9:08 a.m. UTC | #5
I wonder what changed in scarthgap that this issue suddenly appeared?
This does need to be explained before any patches land.

Alex

On Fri, 15 Aug 2025 at 10:09, Peter Marko via lists.openembedded.org
<peter.marko=siemens.com@lists.openembedded.org> wrote:
>
>
>
> > -----Original Message-----
> > From: Marko, Peter (FT D EU SK BFS1)
> > Sent: Friday, August 15, 2025 8:12
> > To: Lau, Karn Jye <karn.jye.lau@intel.com>; openembedded-
> > devel@lists.openembedded.org
> > Subject: RE: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git fetcher
> > conflict by using SRC_URI.
> >
> > Oops, sorry for missing the scarthgap tag.
> > This is good patch, I wonder how this got through the testing.
> >
> > Unfortunately, meta-openembedded LTS branches get updates only once a month
> > so this will be broken trough then next release.
> >
> > Peter
> >
> > > -----Original Message-----
> > > From: Lau, Karn Jye <karn.jye.lau@intel.com>
> > > Sent: Friday, August 15, 2025 8:09
> > > To: Marko, Peter (FT D EU SK BFS1) <Peter.Marko@siemens.com>;
> > > openembedded-devel@lists.openembedded.org
> > > Subject: RE: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git fetcher
> > > conflict by using SRC_URI.
> > >
> > > Hi Peter, it may working fine in master branch
> > > but it not working at all at LTS branch for example yocto 5.0 scargapth, I
> > observing
> > > our daily build hit into this issue.
> > > Thus submitting this patch
> > >
> > > KJ
> > > -----Original Message-----
> > > From: Marko, Peter <Peter.Marko@siemens.com>
> > > Sent: Friday, August 15, 2025 1:53 PM
> > > To: Lau, Karn Jye <karn.jye.lau@intel.com>; openembedded-
> > > devel@lists.openembedded.org
> > > Subject: RE: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git fetcher
> > > conflict by using SRC_URI.
> > >
> > > Both tag and SRVREV works with latest oe-core/poky master.
> > > And is actually encouraged to be used so that hash is verified against the tag.
> > >
> > > I wonder what could be wrong since this is a second patch for this recipe
> > > complaining about errors.
> > >
> > > Peter
> > >
> > > > -----Original Message-----
> > > > From: openembedded-devel@lists.openembedded.org <openembedded-
> > > > devel@lists.openembedded.org> On Behalf Of Lau, Karn Jye via
> > > > lists.openembedded.org
> > > > Sent: Friday, August 15, 2025 6:11
> > > > To: openembedded-devel@lists.openembedded.org
> > > > Subject: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git
> > > > fetcher conflict by using SRC_URI.
> > > >
> > > > From: kjlau0112 <karn.jye.lau@intel.com>
> > > >
> > > > if both SRC_URI & SRCREV specify particular source revision will lead
> > > > to "Fix Fetcher failure: Conflicting revisions error"
> > > >
> > > > Fix by removing SRCREV and specify expected source revision in SRC_URI
> > > > by adding revision parmater.
> > > >
> > > > Signed-off-by: kjlau0112 <karn.jye.lau@intel.com>
> > > > ---
> > > >  meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb | 4
> > > > +---
> > > >  1 file changed, 1 insertion(+), 3 deletions(-)
> > > >
> > > > diff --git
> > > > a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> > > > b/meta- networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> > > > index 88b50a406..2591e2cde 100644
> > > > --- a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> > > > +++ b/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> > > > @@ -23,12 +23,10 @@ LIC_FILES_CHKSUM =
> > > > "file://LICENSE;md5=379d5819937a6c2f1ef1630d341e026d"
> > > >  SECTION = "libs"
> > > >
> > > >  S = "${WORKDIR}/git"
> > > > -SRC_URI = "gitsm://github.com/Mbed-
> > > > TLS/mbedtls.git;protocol=https;branch=mbedtls-3.6;tag=v${PV} \
> > > > +SRC_URI = "gitsm://github.com/Mbed-
> > > > TLS/mbedtls.git;protocol=https;branch=mbedtls-
> > > > 3.6;rev=c765c831e5c2a0971410692f92f7a81d6ec65ec2\
> > > >             file://run-ptest \
> > > >             "
> > > >
> > > > -SRCREV = "c765c831e5c2a0971410692f92f7a81d6ec65ec2"
>
> Could you please keep the SRCREV and just remove the ";tag=v${PV}"?
> Adding rev to SRC_URI is not according to current standards.
>
> Peter
>
> > > > -
> > > >  UPSTREAM_CHECK_GITTAGREGEX = "v(?P<pver>\d+(\.\d+)+)"
> > > >
> > > >  inherit cmake update-alternatives ptest
> > > > --
> > > > 2.34.1
>
>
> -=-=-=-=-=-=-=-=-=-=-=-
> Links: You receive all messages sent to this group.
> View/Reply Online (#118961): https://lists.openembedded.org/g/openembedded-devel/message/118961
> Mute This Topic: https://lists.openembedded.org/mt/114713300/1686489
> Group Owner: openembedded-devel+owner@lists.openembedded.org
> Unsubscribe: https://lists.openembedded.org/g/openembedded-devel/unsub [alex.kanavin@gmail.com]
> -=-=-=-=-=-=-=-=-=-=-=-
>
Lau, Karn Jye Aug. 15, 2025, 9:21 a.m. UTC | #6
Hi Alex, here the changes 

https://github.com/openembedded/meta-openembedded/commit/6dedea42620332987d92044a7e8c8e85fa07c107
when this changes was make, scarthgap  already unbuildable 
unfortunately it get bump further to version 3.6.4 regardless.
https://github.com/openembedded/meta-openembedded/commit/2c9126bd0dd0f62021f5db34d698489bb814ebda

KJ

-----Original Message-----
From: Alexander Kanavin <alex.kanavin@gmail.com> 
Sent: Friday, August 15, 2025 5:08 PM
To: peter.marko@siemens.com
Cc: Lau, Karn Jye <karn.jye.lau@intel.com>; openembedded-devel@lists.openembedded.org
Subject: Re: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git fetcher conflict by using SRC_URI.

I wonder what changed in scarthgap that this issue suddenly appeared?
This does need to be explained before any patches land.

Alex

On Fri, 15 Aug 2025 at 10:09, Peter Marko via lists.openembedded.org <peter.marko=siemens.com@lists.openembedded.org> wrote:
>
>
>
> > -----Original Message-----
> > From: Marko, Peter (FT D EU SK BFS1)
> > Sent: Friday, August 15, 2025 8:12
> > To: Lau, Karn Jye <karn.jye.lau@intel.com>; openembedded- 
> > devel@lists.openembedded.org
> > Subject: RE: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix 
> > git fetcher conflict by using SRC_URI.
> >
> > Oops, sorry for missing the scarthgap tag.
> > This is good patch, I wonder how this got through the testing.
> >
> > Unfortunately, meta-openembedded LTS branches get updates only once 
> > a month so this will be broken trough then next release.
> >
> > Peter
> >
> > > -----Original Message-----
> > > From: Lau, Karn Jye <karn.jye.lau@intel.com>
> > > Sent: Friday, August 15, 2025 8:09
> > > To: Marko, Peter (FT D EU SK BFS1) <Peter.Marko@siemens.com>; 
> > > openembedded-devel@lists.openembedded.org
> > > Subject: RE: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix 
> > > git fetcher conflict by using SRC_URI.
> > >
> > > Hi Peter, it may working fine in master branch but it not working 
> > > at all at LTS branch for example yocto 5.0 scargapth, I
> > observing
> > > our daily build hit into this issue.
> > > Thus submitting this patch
> > >
> > > KJ
> > > -----Original Message-----
> > > From: Marko, Peter <Peter.Marko@siemens.com>
> > > Sent: Friday, August 15, 2025 1:53 PM
> > > To: Lau, Karn Jye <karn.jye.lau@intel.com>; openembedded- 
> > > devel@lists.openembedded.org
> > > Subject: RE: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix 
> > > git fetcher conflict by using SRC_URI.
> > >
> > > Both tag and SRVREV works with latest oe-core/poky master.
> > > And is actually encouraged to be used so that hash is verified against the tag.
> > >
> > > I wonder what could be wrong since this is a second patch for this 
> > > recipe complaining about errors.
> > >
> > > Peter
> > >
> > > > -----Original Message-----
> > > > From: openembedded-devel@lists.openembedded.org <openembedded- 
> > > > devel@lists.openembedded.org> On Behalf Of Lau, Karn Jye via 
> > > > lists.openembedded.org
> > > > Sent: Friday, August 15, 2025 6:11
> > > > To: openembedded-devel@lists.openembedded.org
> > > > Subject: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix 
> > > > git fetcher conflict by using SRC_URI.
> > > >
> > > > From: kjlau0112 <karn.jye.lau@intel.com>
> > > >
> > > > if both SRC_URI & SRCREV specify particular source revision will 
> > > > lead to "Fix Fetcher failure: Conflicting revisions error"
> > > >
> > > > Fix by removing SRCREV and specify expected source revision in 
> > > > SRC_URI by adding revision parmater.
> > > >
> > > > Signed-off-by: kjlau0112 <karn.jye.lau@intel.com>
> > > > ---
> > > >  meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb | 
> > > > 4
> > > > +---
> > > >  1 file changed, 1 insertion(+), 3 deletions(-)
> > > >
> > > > diff --git
> > > > a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> > > > b/meta- networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> > > > index 88b50a406..2591e2cde 100644
> > > > --- 
> > > > a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
> > > > +++ b/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4
> > > > +++ .bb
> > > > @@ -23,12 +23,10 @@ LIC_FILES_CHKSUM = 
> > > > "file://LICENSE;md5=379d5819937a6c2f1ef1630d341e026d"
> > > >  SECTION = "libs"
> > > >
> > > >  S = "${WORKDIR}/git"
> > > > -SRC_URI = "gitsm://github.com/Mbed- 
> > > > TLS/mbedtls.git;protocol=https;branch=mbedtls-3.6;tag=v${PV} \
> > > > +SRC_URI = "gitsm://github.com/Mbed-
> > > > TLS/mbedtls.git;protocol=https;branch=mbedtls-
> > > > 3.6;rev=c765c831e5c2a0971410692f92f7a81d6ec65ec2\
> > > >             file://run-ptest \
> > > >             "
> > > >
> > > > -SRCREV = "c765c831e5c2a0971410692f92f7a81d6ec65ec2"
>
> Could you please keep the SRCREV and just remove the ";tag=v${PV}"?
> Adding rev to SRC_URI is not according to current standards.
>
> Peter
>
> > > > -
> > > >  UPSTREAM_CHECK_GITTAGREGEX = "v(?P<pver>\d+(\.\d+)+)"
> > > >
> > > >  inherit cmake update-alternatives ptest
> > > > --
> > > > 2.34.1
>
>
> -=-=-=-=-=-=-=-=-=-=-=-
> Links: You receive all messages sent to this group.
> View/Reply Online (#118961): 
> https://lists.openembedded.org/g/openembedded-devel/message/118961
> Mute This Topic: https://lists.openembedded.org/mt/114713300/1686489
> Group Owner: openembedded-devel+owner@lists.openembedded.org
> Unsubscribe: https://lists.openembedded.org/g/openembedded-devel/unsub 
> [alex.kanavin@gmail.com]
> -=-=-=-=-=-=-=-=-=-=-=-
>
Alexander Kanavin Aug. 15, 2025, 9:29 a.m. UTC | #7
On Fri, 15 Aug 2025 at 11:21, Lau, Karn Jye <karn.jye.lau@intel.com> wrote:
> https://github.com/openembedded/meta-openembedded/commit/6dedea42620332987d92044a7e8c8e85fa07c107
> when this changes was make, scarthgap  already unbuildable
> unfortunately it get bump further to version 3.6.4 regardless.
> https://github.com/openembedded/meta-openembedded/commit/2c9126bd0dd0f62021f5db34d698489bb814ebda

But how did either change manage to land in scarthgap without failing
in testing at any point? Something doesn't compute here, especially
because people started reporting the failure only today.

Alex
Gyorgy Sarvari Aug. 15, 2025, 9:34 a.m. UTC | #8
On 8/15/25 11:29, Alexander Kanavin via lists.openembedded.org wrote:
> On Fri, 15 Aug 2025 at 11:21, Lau, Karn Jye <karn.jye.lau@intel.com> wrote:
>> https://github.com/openembedded/meta-openembedded/commit/6dedea42620332987d92044a7e8c8e85fa07c107
>> when this changes was make, scarthgap  already unbuildable
>> unfortunately it get bump further to version 3.6.4 regardless.
>> https://github.com/openembedded/meta-openembedded/commit/2c9126bd0dd0f62021f5db34d698489bb814ebda
> But how did either change manage to land in scarthgap without failing
> in testing at any point? Something doesn't compute here, especially
> because people started reporting the failure only today.

I can only answer the last part: the merge happened yesterday evening,
that's why the reports are coming now.

>
> Alex
>
> -=-=-=-=-=-=-=-=-=-=-=-
> Links: You receive all messages sent to this group.
> View/Reply Online (#118966): https://lists.openembedded.org/g/openembedded-devel/message/118966
> Mute This Topic: https://lists.openembedded.org/mt/114713300/6084445
> Group Owner: openembedded-devel+owner@lists.openembedded.org
> Unsubscribe: https://lists.openembedded.org/g/openembedded-devel/unsub [skandigraun@gmail.com]
> -=-=-=-=-=-=-=-=-=-=-=-
>
Peter Marko Aug. 15, 2025, 9:36 a.m. UTC | #9
> -----Original Message-----
> From: Alexander Kanavin <alex.kanavin@gmail.com>
> Sent: Friday, August 15, 2025 11:30
> To: Lau, Karn Jye <karn.jye.lau@intel.com>
> Cc: Marko, Peter (FT D EU SK BFS1) <Peter.Marko@siemens.com>;
> openembedded-devel@lists.openembedded.org
> Subject: Re: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git fetcher
> conflict by using SRC_URI.
> 
> On Fri, 15 Aug 2025 at 11:21, Lau, Karn Jye <karn.jye.lau@intel.com> wrote:
> > https://github.com/openembedded/meta-
> openembedded/commit/6dedea42620332987d92044a7e8c8e85fa07c107
> > when this changes was make, scarthgap  already unbuildable
> > unfortunately it get bump further to version 3.6.4 regardless.
> > https://github.com/openembedded/meta-
> openembedded/commit/2c9126bd0dd0f62021f5db34d698489bb814ebda
> 
> But how did either change manage to land in scarthgap without failing
> in testing at any point? Something doesn't compute here, especially
> because people started reporting the failure only today.

This was moved from scathgap-next to scarthgap yesterday, so timing is clear
https://lists.openembedded.org/g/openembedded-devel/message/118942

The successful autobuilder runs are a mystery.
Maybe the code throws only if the sources are not in downloads (e.g. autobuilder has these from master).

Peter

> 
> Alex
Alexander Kanavin Aug. 15, 2025, 9:39 a.m. UTC | #10
On Fri, 15 Aug 2025 at 11:36, Marko, Peter <Peter.Marko@siemens.com> wrote:
> This was moved from scathgap-next to scarthgap yesterday, so timing is clear
> https://lists.openembedded.org/g/openembedded-devel/message/118942
>
> The successful autobuilder runs are a mystery.
> Maybe the code throws only if the sources are not in downloads (e.g. autobuilder has these from master).

The more worrying scenario is that there was no actual build at any
point before this landed, I hope that is not the case.

I don't object to the fixes, just want to find out where the quality
gating process failed.

Alex
Mikko Rapeli Aug. 15, 2025, 9:53 a.m. UTC | #11
Hi,

On Fri, Aug 15, 2025 at 11:39:25AM +0200, Alexander Kanavin via lists.openembedded.org wrote:
> On Fri, 15 Aug 2025 at 11:36, Marko, Peter <Peter.Marko@siemens.com> wrote:
> > This was moved from scathgap-next to scarthgap yesterday, so timing is clear
> > https://lists.openembedded.org/g/openembedded-devel/message/118942
> >
> > The successful autobuilder runs are a mystery.
> > Maybe the code throws only if the sources are not in downloads (e.g. autobuilder has these from master).
> 
> The more worrying scenario is that there was no actual build at any
> point before this landed, I hope that is not the case.
> 
> I don't object to the fixes, just want to find out where the quality
> gating process failed.

FWIW, most users of mbedtls in the Arm firmware land embed the git repo or sources
directly and don't use this recipe. This is problematic for SBOM, CVE security
handling etc but this how things are currently. Users are u-boot, trusted-firmware-a,
trusted-firmware-m, OP-TEE etc. Thus this recipe may have very few users even when
mbedtls is used in a lot of places.

Cheers,

-Mikko
Lau, Karn Jye Aug. 15, 2025, 11:05 a.m. UTC | #12
Hi folks I resubmitted a new patch , just realized that will be better instead of trying to reply the same thread, again  I apologize for unintentionally "spamming"
I try the best to contribute, upstream fix and contribution is not the best of my knowledge, I will try improve at my end 

KJ

-----Original Message-----
From: Mikko Rapeli <mikko.rapeli@linaro.org> 
Sent: Friday, August 15, 2025 5:53 PM
To: alex.kanavin@gmail.com
Cc: Marko, Peter <Peter.Marko@siemens.com>; Lau, Karn Jye <karn.jye.lau@intel.com>; openembedded-devel@lists.openembedded.org
Subject: Re: [oe] [meta-networking][scarthgap][PATCH] mbedtls: fix git fetcher conflict by using SRC_URI.

Hi,

On Fri, Aug 15, 2025 at 11:39:25AM +0200, Alexander Kanavin via lists.openembedded.org wrote:
> On Fri, 15 Aug 2025 at 11:36, Marko, Peter <Peter.Marko@siemens.com> wrote:
> > This was moved from scathgap-next to scarthgap yesterday, so timing 
> > is clear
> > https://lists.openembedded.org/g/openembedded-devel/message/118942
> >
> > The successful autobuilder runs are a mystery.
> > Maybe the code throws only if the sources are not in downloads (e.g. autobuilder has these from master).
> 
> The more worrying scenario is that there was no actual build at any 
> point before this landed, I hope that is not the case.
> 
> I don't object to the fixes, just want to find out where the quality 
> gating process failed.

FWIW, most users of mbedtls in the Arm firmware land embed the git repo or sources directly and don't use this recipe. This is problematic for SBOM, CVE security handling etc but this how things are currently. Users are u-boot, trusted-firmware-a, trusted-firmware-m, OP-TEE etc. Thus this recipe may have very few users even when mbedtls is used in a lot of places.

Cheers,

-Mikko
From: kjlau0112 <karn.jye.lau@intel.com>

Keep SRCREV and remove tag paramater to avoid
Fix Fetcher failure: Conflicting revisions error

Signed-off-by: kjlau0112 <karn.jye.lau@intel.com>
---
 meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb b/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
index 88b50a406..91adf6f5a 100644
--- a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
+++ b/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
@@ -23,7 +23,7 @@ LIC_FILES_CHKSUM = "file://LICENSE;md5=379d5819937a6c2f1ef1630d341e026d"
 SECTION = "libs"

 S = "${WORKDIR}/git"
-SRC_URI = "gitsm://github.com/Mbed-TLS/mbedtls.git;protocol=https;branch=mbedtls-3.6;tag=v${PV} \
+SRC_URI = "gitsm://github.com/Mbed-TLS/mbedtls.git;protocol=https;branch=mbedtls-3.6;\
            file://run-ptest \
            "

--
2.34.1
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#118974): https://lists.openembedded.org/g/openembedded-devel/message/118974
Mute This Topic: https://lists.openembedded.org/mt/114715887/6860922
Group Owner: openembedded-devel+owner@lists.openembedded.org
Unsubscribe: https://lists.openembedded.org/g/openembedded-devel/unsub [karn.jye.lau@intel.com]
-=-=-=-=-=-=-=-=-=-=-=-
diff mbox series

Patch

diff --git a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb b/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
index 88b50a406..2591e2cde 100644
--- a/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
+++ b/meta-networking/recipes-connectivity/mbedtls/mbedtls_3.6.4.bb
@@ -23,12 +23,10 @@  LIC_FILES_CHKSUM = "file://LICENSE;md5=379d5819937a6c2f1ef1630d341e026d"
 SECTION = "libs"
 
 S = "${WORKDIR}/git"
-SRC_URI = "gitsm://github.com/Mbed-TLS/mbedtls.git;protocol=https;branch=mbedtls-3.6;tag=v${PV} \
+SRC_URI = "gitsm://github.com/Mbed-TLS/mbedtls.git;protocol=https;branch=mbedtls-3.6;rev=c765c831e5c2a0971410692f92f7a81d6ec65ec2\
            file://run-ptest \
            "
 
-SRCREV = "c765c831e5c2a0971410692f92f7a81d6ec65ec2"
-
 UPSTREAM_CHECK_GITTAGREGEX = "v(?P<pver>\d+(\.\d+)+)"
 
 inherit cmake update-alternatives ptest