[PATCH 1/3] go: drop patch to make content based hash generation less pedan

[PATCH 1/3] go: drop patch to make content based hash generation less pedan

[Jose Quaresma]

With patch in place I have see some build issues mainly in meta-virtualization
with ti am62xx-evm, am64xx-evm and riscv64 machines.

All the mahines fails to build the following recipes:

| Summary: 5 tasks failed:
|  /srv/oe/build/conf/../../layers/meta-virtualization/recipes-containers/runc/runc-opencontainers_git.bb:do_compile
|  /srv/oe/build/conf/../../layers/meta-virtualization/recipes-containers/skopeo/skopeo_git.bb:do_compile
|  /srv/oe/build/conf/../../layers/meta-lmp/meta-lmp-base/recipes-support/fioconfig/fioconfig_git.bb:do_compile
|  /srv/oe/build/conf/../../layers/meta-virtualization/recipes-containers/docker/docker-ce_git.bb:do_compile
|  /srv/oe/build/conf/../../layers/meta-virtualization/recipes-containers/containerd/containerd-opencontainers_git.bb:do_compile
| Summary: There were 8 ERROR messages, returning a non-zero exit code.

And all of them fails with the same issue:

| panic: runtime error: index out of range [0] with length 0
|
| goroutine 6688 [running]:
| cmd/go/internal/work.(*Builder).buildActionID(0xc0000030e0, 0xc000172a00)
| 	/srv/oe/build/tmp-lmp/work/x86_64-linux/go-cross-riscv64/1.19-r0/recipe-sysroot-native/usr/lib/riscv64-lmp-linux/go/src/cmd/go/internal/work/exec.go:281 +0x2425
| cmd/go/internal/work.(*Builder).build(0xc0000030e0, {0xc000000238?, 0x0?}, 0xc000172a00)
| 	/srv/oe/build/tmp-lmp/work/x86_64-linux/go-cross-riscv64/1.19-r0/recipe-sysroot-native/usr/lib/riscv64-lmp-linux/go/src/cmd/go/internal/work/exec.go:478 +0x199
| cmd/go/internal/work.(*Builder).Do.func2({0xb2e4f0, 0xc000000238}, 0xc000172a00)
| 	/srv/oe/build/tmp-lmp/work/x86_64-linux/go-cross-riscv64/1.19-r0/recipe-sysroot-native/usr/lib/riscv64-lmp-linux/go/src/cmd/go/internal/work/exec.go:139 +0x5fc
| cmd/go/internal/work.(*Builder).Do.func3()
| 	/srv/oe/build/tmp-lmp/work/x86_64-linux/go-cross-riscv64/1.19-r0/recipe-sysroot-native/usr/lib/riscv64-lmp-linux/go/src/cmd/go/internal/work/exec.go:201 +0xb9
| created by cmd/go/internal/work.(*Builder).Do
| 	/srv/oe/build/tmp-lmp/work/x86_64-linux/go-cross-riscv64/1.19-r0/recipe-sysroot-native/usr/lib/riscv64-lmp-linux/go/src/cmd/go/internal/work/exec.go:187 +0x3f5
| make: *** [Makefile:244: bin/containerd] Error 2
| make: *** Waiting for unfinished jobs....

Without this patch everithing look good.

Signed-off-by: Jose Quaresma <jose.quaresma@foundries.io>
---
 meta/recipes-devtools/go/go-1.19.inc | 1 -
 1 file changed, 1 deletion(-)

diff --git a/meta/recipes-devtools/go/go-1.19.inc b/meta/recipes-devtools/go/go-1.19.inc
index f733a807b4..355a82eb9f 100644
--- a/meta/recipes-devtools/go/go-1.19.inc
+++ b/meta/recipes-devtools/go/go-1.19.inc
@@ -5,7 +5,6 @@ FILESEXTRAPATHS:prepend := "${FILE_DIRNAME}/go:"
 LIC_FILES_CHKSUM = "file://LICENSE;md5=5d4950ecb7b26d2c5e4e7b4e0dd74707"
 
 SRC_URI += "\
-    file://0001-cmd-go-make-content-based-hash-generation-less-pedan.patch \
     file://0003-allow-GOTOOLDIR-to-be-overridden-in-the-environment.patch \
     file://0004-ld-add-soname-to-shareable-objects.patch \
     file://0005-make.bash-override-CC-when-building-dist-and-go_boot.patch \
-- 
2.34.1

[PATCH 2/3] kern-tools-native: don't pull the toolchain as it is not used

[Jose Quaresma]

Prevents the default dependencies, namely the C compiler and standard C library (libc),
from being added to DEPENDS. This variable is usually used within recipes that do not
require any compilation using the C compiler.

This also improves the speed as it will not populate the
native sysroot with the toolchain.

Signed-off-by: Jose Quaresma <jose.quaresma@foundries.io>
---
 meta/recipes-kernel/kern-tools/kern-tools-native_git.bb | 9 +++------
 1 file changed, 3 insertions(+), 6 deletions(-)

diff --git a/meta/recipes-kernel/kern-tools/kern-tools-native_git.bb b/meta/recipes-kernel/kern-tools/kern-tools-native_git.bb
index dea7b65a7c..5fe46341da 100644
--- a/meta/recipes-kernel/kern-tools/kern-tools-native_git.bb
+++ b/meta/recipes-kernel/kern-tools/kern-tools-native_git.bb
@@ -19,13 +19,10 @@ inherit native
 SRC_URI = "git://git.yoctoproject.org/yocto-kernel-tools.git;branch=master"
 S = "${WORKDIR}/git"
 
-do_configure() {
-	:
-}
+INHIBIT_DEFAULT_DEPS = "1"
 
-do_compile() { 
-	:
-}
+do_configure[noexec] = "1"
+do_compile[noexec] = "1"
 
 do_install() {
 	oe_runmake DESTDIR=${D}${bindir} install
-- 
2.34.1

Re: [OE-core] [PATCH 2/3] kern-tools-native: don't pull the toolchain as it is not used

[Ross Burton]

On 30 Sep 2022, at 18:54, Jose Quaresma via lists.openembedded.org <quaresma.jose=gmail.com@lists.openembedded.org> wrote:
> 
> Prevents the default dependencies, namely the C compiler and standard C library (libc),
> from being added to DEPENDS. This variable is usually used within recipes that do not
> require any compilation using the C compiler.
> 
> This also improves the speed as it will not populate the
> native sysroot with the toolchain.

But this is a native recipe, so it doesn’t depend on a cross-compiler:

$ bitbake-getvar -r kern-tools-native DEPENDS
DEPENDS=" git-native”

Ross

Re: [OE-core] [PATCH 2/3] kern-tools-native: don't pull the toolchain as it is not used

[Jose Quaresma]

[-- Attachment #1: Type: text/plain, Size: 985 bytes --]

Hi Ross,

Ross Burton <Ross.Burton@arm.com> escreveu no dia quinta, 6/10/2022 à(s)
11:50:

> On 30 Sep 2022, at 18:54, Jose Quaresma via lists.openembedded.org
> <quaresma.jose=gmail.com@lists.openembedded.org> wrote:
> >
> > Prevents the default dependencies, namely the C compiler and standard C
> library (libc),
> > from being added to DEPENDS. This variable is usually used within
> recipes that do not
> > require any compilation using the C compiler.
> >
> > This also improves the speed as it will not populate the
> > native sysroot with the toolchain.
>
> But this is a native recipe, so it doesn’t depend on a cross-compiler:
>
> $ bitbake-getvar -r kern-tools-native DEPENDS
> DEPENDS=" git-native”
>

Right, sorry for my confusion that maybe it was ignorance of the reality of
this detail of native packages.
Please drop that patch and thanks for your time in reviewing it.

Jose


>
> Ross



-- 
Best regards,

José Quaresma

[-- Attachment #2: Type: text/html, Size: 1739 bytes --]

Re: [OE-core] [PATCH 2/3] kern-tools-native: don't pull the toolchain as it is not used

[Jose Quaresma]

[-- Attachment #1: Type: text/plain, Size: 1855 bytes --]

Hi Alexandre,

I see this one is still on staging abelloni main-next
https://git.yoctoproject.org/poky-contrib/log/?h=abelloni/master-next

Can this be dropped please?

Jose

Jose Quaresma via lists.openembedded.org <quaresma.jose=
gmail.com@lists.openembedded.org> escreveu no dia sexta, 7/10/2022 à(s)
09:59:

> Hi Ross,
>
> Ross Burton <Ross.Burton@arm.com> escreveu no dia quinta, 6/10/2022 à(s)
> 11:50:
>
>> On 30 Sep 2022, at 18:54, Jose Quaresma via lists.openembedded.org
>> <quaresma.jose=gmail.com@lists.openembedded.org> wrote:
>> >
>> > Prevents the default dependencies, namely the C compiler and standard C
>> library (libc),
>> > from being added to DEPENDS. This variable is usually used within
>> recipes that do not
>> > require any compilation using the C compiler.
>> >
>> > This also improves the speed as it will not populate the
>> > native sysroot with the toolchain.
>>
>> But this is a native recipe, so it doesn’t depend on a cross-compiler:
>>
>> $ bitbake-getvar -r kern-tools-native DEPENDS
>> DEPENDS=" git-native”
>>
>
> Right, sorry for my confusion that maybe it was ignorance of the reality
> of this detail of native packages.
> Please drop that patch and thanks for your time in reviewing it.
>
> Jose
>
>
>>
>> Ross
>
>
>
> --
> Best regards,
>
> José Quaresma
>
> -=-=-=-=-=-=-=-=-=-=-=-
> Links: You receive all messages sent to this group.
> View/Reply Online (#171519):
> https://lists.openembedded.org/g/openembedded-core/message/171519
> Mute This Topic: https://lists.openembedded.org/mt/94022665/5052612
> Group Owner: openembedded-core+owner@lists.openembedded.org
> Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [
> quaresma.jose@gmail.com]
> -=-=-=-=-=-=-=-=-=-=-=-
>
>

-- 
Best regards,

José Quaresma

[-- Attachment #2: Type: text/html, Size: 3590 bytes --]

Re: [OE-core] [PATCH 2/3] kern-tools-native: don't pull the toolchain as it is not used

[Alexandre Belloni]

On 24/10/2022 15:21:56+0100, Jose Quaresma wrote:
> Hi Alexandre,
> 
> I see this one is still on staging abelloni main-next
> https://git.yoctoproject.org/poky-contrib/log/?h=abelloni/master-next
> 
> Can this be dropped please?
> 

Dropped now

> Jose
> 
> Jose Quaresma via lists.openembedded.org <quaresma.jose=
> gmail.com@lists.openembedded.org> escreveu no dia sexta, 7/10/2022 à(s)
> 09:59:
> 
> > Hi Ross,
> >
> > Ross Burton <Ross.Burton@arm.com> escreveu no dia quinta, 6/10/2022 à(s)
> > 11:50:
> >
> >> On 30 Sep 2022, at 18:54, Jose Quaresma via lists.openembedded.org
> >> <quaresma.jose=gmail.com@lists.openembedded.org> wrote:
> >> >
> >> > Prevents the default dependencies, namely the C compiler and standard C
> >> library (libc),
> >> > from being added to DEPENDS. This variable is usually used within
> >> recipes that do not
> >> > require any compilation using the C compiler.
> >> >
> >> > This also improves the speed as it will not populate the
> >> > native sysroot with the toolchain.
> >>
> >> But this is a native recipe, so it doesn’t depend on a cross-compiler:
> >>
> >> $ bitbake-getvar -r kern-tools-native DEPENDS
> >> DEPENDS=" git-native”
> >>
> >
> > Right, sorry for my confusion that maybe it was ignorance of the reality
> > of this detail of native packages.
> > Please drop that patch and thanks for your time in reviewing it.
> >
> > Jose
> >
> >
> >>
> >> Ross
> >
> >
> >
> > --
> > Best regards,
> >
> > José Quaresma
> >
> > -=-=-=-=-=-=-=-=-=-=-=-
> > Links: You receive all messages sent to this group.
> > View/Reply Online (#171519):
> > https://lists.openembedded.org/g/openembedded-core/message/171519
> > Mute This Topic: https://lists.openembedded.org/mt/94022665/5052612
> > Group Owner: openembedded-core+owner@lists.openembedded.org
> > Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [
> > quaresma.jose@gmail.com]
> > -=-=-=-=-=-=-=-=-=-=-=-
> >
> >
> 
> -- 
> Best regards,
> 
> José Quaresma

-- 
Alexandre Belloni, co-owner and COO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com

[PATCH 3/3] kernel-yocto: improve fatal error messages of symbol_why.py

[Jose Quaresma]

Improve the fatal error message of the yocto-kernel-tools symbol_why.py
and shows the command that generate the error as it can help understand
the root cause of the error.

Signed-off-by: Jose Quaresma <jose.quaresma@foundries.io>
---
 meta/classes-recipe/kernel-yocto.bbclass | 15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)

diff --git a/meta/classes-recipe/kernel-yocto.bbclass b/meta/classes-recipe/kernel-yocto.bbclass
index 8eda0dcaf3..46df9b23dc 100644
--- a/meta/classes-recipe/kernel-yocto.bbclass
+++ b/meta/classes-recipe/kernel-yocto.bbclass
@@ -573,9 +573,10 @@ python do_kernel_configcheck() {
 
     # category #1: mismatches
     try:
-        analysis = subprocess.check_output(['symbol_why.py', '--dotconfig',  '{}'.format( d.getVar('B') + '/.config' ), '--mismatches', extra_params], cwd=s, env=env ).decode('utf-8')
+        cmd = ['symbol_why.py', '--dotconfig',  '{}'.format( d.getVar('B') + '/.config' ), '--mismatches', extra_params]
+        analysis = subprocess.check_output(cmd, cwd=s, env=env ).decode('utf-8')
     except subprocess.CalledProcessError as e:
-        bb.fatal( "config analysis failed: %s" % e.output.decode('utf-8'))
+        bb.fatal( "config analysis failed when running '%s': %s" % (" ".join(cmd), e.output.decode('utf-8')))
 
     if analysis:
         outfile = "{}/{}/cfg/mismatch.txt".format( s, kmeta )
@@ -595,9 +596,10 @@ python do_kernel_configcheck() {
     if bsp_check_visibility > 1:
         extra_params = "--strict"
     try:
-        analysis = subprocess.check_output(['symbol_why.py', '--dotconfig',  '{}'.format( d.getVar('B') + '/.config' ), '--invalid', extra_params], cwd=s, env=env ).decode('utf-8')
+        cmd = ['symbol_why.py', '--dotconfig',  '{}'.format( d.getVar('B') + '/.config' ), '--invalid', extra_params]
+        analysis = subprocess.check_output(cmd, cwd=s, env=env ).decode('utf-8')
     except subprocess.CalledProcessError as e:
-        bb.fatal( "config analysis failed: %s" % e.output.decode('utf-8'))
+        bb.fatal( "config analysis failed when running '%s': %s" % (" ".join(cmd), e.output.decode('utf-8')))
 
     if analysis:
         outfile = "{}/{}/cfg/invalid.txt".format(s,kmeta)
@@ -614,9 +616,10 @@ python do_kernel_configcheck() {
 
     # category #3: redefined options (this is pretty verbose and is debug only)
     try:
-        analysis = subprocess.check_output(['symbol_why.py', '--dotconfig',  '{}'.format( d.getVar('B') + '/.config' ), '--sanity'], cwd=s, env=env ).decode('utf-8')
+        cmd = ['symbol_why.py', '--dotconfig',  '{}'.format( d.getVar('B') + '/.config' ), '--sanity']
+        analysis = subprocess.check_output(cmd, cwd=s, env=env ).decode('utf-8')
     except subprocess.CalledProcessError as e:
-        bb.fatal( "config analysis failed: %s" % e.output.decode('utf-8'))
+        bb.fatal( "config analysis failed when running '%s': %s" % (" ".join(cmd), e.output.decode('utf-8')))
 
     if analysis:
         outfile = "{}/{}/cfg/redefinition.txt".format(s,kmeta)
-- 
2.34.1

Re: [OE-core] [PATCH 3/3] kernel-yocto: improve fatal error messages of symbol_why.py

[Ross Burton]

> On 30 Sep 2022, at 18:54, Jose Quaresma via lists.openembedded.org <quaresma.jose=gmail.com@lists.openembedded.org> wrote:
> -        bb.fatal( "config analysis failed: %s" % e.output.decode('utf-8'))
> +        bb.fatal( "config analysis failed when running '%s': %s" % (" ".join(cmd), e.output.decode('utf-8')))

Much easier to just use e.cmd.

https://docs.python.org/3/library/subprocess.html#subprocess.CalledProcessError

Ross

Re: [OE-core] [PATCH 3/3] kernel-yocto: improve fatal error messages of symbol_why.py

[Jose Quaresma]

[-- Attachment #1: Type: text/plain, Size: 678 bytes --]

Ross Burton <Ross.Burton@arm.com> escreveu no dia quinta, 6/10/2022 à(s)
11:51:

>
>
> > On 30 Sep 2022, at 18:54, Jose Quaresma via lists.openembedded.org
> <quaresma.jose=gmail.com@lists.openembedded.org> wrote:
> > -        bb.fatal( "config analysis failed: %s" %
> e.output.decode('utf-8'))
> > +        bb.fatal( "config analysis failed when running '%s': %s" % ("
> ".join(cmd), e.output.decode('utf-8')))
>
> Much easier to just use e.cmd.
>

It's really much simpler. thank you, I will send a v2

Jose


>
>
> https://docs.python.org/3/library/subprocess.html#subprocess.CalledProcessError
>
> Ross



-- 
Best regards,

José Quaresma

[-- Attachment #2: Type: text/html, Size: 1613 bytes --]

Re: [OE-core] [PATCH 1/3] go: drop patch to make content based hash generation less pedan

[Alexander Kanavin]

You need to remove the actual patch, and run the reproducibility test.

Alex

On Fri, 30 Sept 2022 at 19:55, Jose Quaresma <quaresma.jose@gmail.com> wrote:
>
> With patch in place I have see some build issues mainly in meta-virtualization
> with ti am62xx-evm, am64xx-evm and riscv64 machines.
>
> All the mahines fails to build the following recipes:
>
> | Summary: 5 tasks failed:
> |  /srv/oe/build/conf/../../layers/meta-virtualization/recipes-containers/runc/runc-opencontainers_git.bb:do_compile
> |  /srv/oe/build/conf/../../layers/meta-virtualization/recipes-containers/skopeo/skopeo_git.bb:do_compile
> |  /srv/oe/build/conf/../../layers/meta-lmp/meta-lmp-base/recipes-support/fioconfig/fioconfig_git.bb:do_compile
> |  /srv/oe/build/conf/../../layers/meta-virtualization/recipes-containers/docker/docker-ce_git.bb:do_compile
> |  /srv/oe/build/conf/../../layers/meta-virtualization/recipes-containers/containerd/containerd-opencontainers_git.bb:do_compile
> | Summary: There were 8 ERROR messages, returning a non-zero exit code.
>
> And all of them fails with the same issue:
>
> | panic: runtime error: index out of range [0] with length 0
> |
> | goroutine 6688 [running]:
> | cmd/go/internal/work.(*Builder).buildActionID(0xc0000030e0, 0xc000172a00)
> |       /srv/oe/build/tmp-lmp/work/x86_64-linux/go-cross-riscv64/1.19-r0/recipe-sysroot-native/usr/lib/riscv64-lmp-linux/go/src/cmd/go/internal/work/exec.go:281 +0x2425
> | cmd/go/internal/work.(*Builder).build(0xc0000030e0, {0xc000000238?, 0x0?}, 0xc000172a00)
> |       /srv/oe/build/tmp-lmp/work/x86_64-linux/go-cross-riscv64/1.19-r0/recipe-sysroot-native/usr/lib/riscv64-lmp-linux/go/src/cmd/go/internal/work/exec.go:478 +0x199
> | cmd/go/internal/work.(*Builder).Do.func2({0xb2e4f0, 0xc000000238}, 0xc000172a00)
> |       /srv/oe/build/tmp-lmp/work/x86_64-linux/go-cross-riscv64/1.19-r0/recipe-sysroot-native/usr/lib/riscv64-lmp-linux/go/src/cmd/go/internal/work/exec.go:139 +0x5fc
> | cmd/go/internal/work.(*Builder).Do.func3()
> |       /srv/oe/build/tmp-lmp/work/x86_64-linux/go-cross-riscv64/1.19-r0/recipe-sysroot-native/usr/lib/riscv64-lmp-linux/go/src/cmd/go/internal/work/exec.go:201 +0xb9
> | created by cmd/go/internal/work.(*Builder).Do
> |       /srv/oe/build/tmp-lmp/work/x86_64-linux/go-cross-riscv64/1.19-r0/recipe-sysroot-native/usr/lib/riscv64-lmp-linux/go/src/cmd/go/internal/work/exec.go:187 +0x3f5
> | make: *** [Makefile:244: bin/containerd] Error 2
> | make: *** Waiting for unfinished jobs....
>
> Without this patch everithing look good.
>
> Signed-off-by: Jose Quaresma <jose.quaresma@foundries.io>
> ---
>  meta/recipes-devtools/go/go-1.19.inc | 1 -
>  1 file changed, 1 deletion(-)
>
> diff --git a/meta/recipes-devtools/go/go-1.19.inc b/meta/recipes-devtools/go/go-1.19.inc
> index f733a807b4..355a82eb9f 100644
> --- a/meta/recipes-devtools/go/go-1.19.inc
> +++ b/meta/recipes-devtools/go/go-1.19.inc
> @@ -5,7 +5,6 @@ FILESEXTRAPATHS:prepend := "${FILE_DIRNAME}/go:"
>  LIC_FILES_CHKSUM = "file://LICENSE;md5=5d4950ecb7b26d2c5e4e7b4e0dd74707"
>
>  SRC_URI += "\
> -    file://0001-cmd-go-make-content-based-hash-generation-less-pedan.patch \
>      file://0003-allow-GOTOOLDIR-to-be-overridden-in-the-environment.patch \
>      file://0004-ld-add-soname-to-shareable-objects.patch \
>      file://0005-make.bash-override-CC-when-building-dist-and-go_boot.patch \
> --
> 2.34.1
>
>
> -=-=-=-=-=-=-=-=-=-=-=-
> Links: You receive all messages sent to this group.
> View/Reply Online (#171233): https://lists.openembedded.org/g/openembedded-core/message/171233
> Mute This Topic: https://lists.openembedded.org/mt/94022663/1686489
> Group Owner: openembedded-core+owner@lists.openembedded.org
> Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [alex.kanavin@gmail.com]
> -=-=-=-=-=-=-=-=-=-=-=-
>

Re: [OE-core] [PATCH 1/3] go: drop patch to make content based hash generation less pedan

[Jose Quaresma]

[-- Attachment #1: Type: text/plain, Size: 4325 bytes --]

Hi Alex,

Alexander Kanavin <alex.kanavin@gmail.com> escreveu no dia sexta,
30/09/2022 à(s) 18:57:

> You need to remove the actual patch, and run the reproducibility test.
>

Forgot to remove the patch! Will send a v2 but I can't run the
reproducibility now.

Jose


> Alex
>
> On Fri, 30 Sept 2022 at 19:55, Jose Quaresma <quaresma.jose@gmail.com>
> wrote:
> >
> > With patch in place I have see some build issues mainly in
> meta-virtualization
> > with ti am62xx-evm, am64xx-evm and riscv64 machines.
> >
> > All the mahines fails to build the following recipes:
> >
> > | Summary: 5 tasks failed:
> > |
> /srv/oe/build/conf/../../layers/meta-virtualization/recipes-containers/runc/runc-opencontainers_git.bb:
> do_compile
> > |
> /srv/oe/build/conf/../../layers/meta-virtualization/recipes-containers/skopeo/skopeo_git.bb:
> do_compile
> > |
> /srv/oe/build/conf/../../layers/meta-lmp/meta-lmp-base/recipes-support/fioconfig/fioconfig_git.bb:
> do_compile
> > |
> /srv/oe/build/conf/../../layers/meta-virtualization/recipes-containers/docker/docker-ce_git.bb:
> do_compile
> > |
> /srv/oe/build/conf/../../layers/meta-virtualization/recipes-containers/containerd/containerd-opencontainers_git.bb:
> do_compile
> > | Summary: There were 8 ERROR messages, returning a non-zero exit code.
> >
> > And all of them fails with the same issue:
> >
> > | panic: runtime error: index out of range [0] with length 0
> > |
> > | goroutine 6688 [running]:
> > | cmd/go/internal/work.(*Builder).buildActionID(0xc0000030e0,
> 0xc000172a00)
> > |
>  /srv/oe/build/tmp-lmp/work/x86_64-linux/go-cross-riscv64/1.19-r0/recipe-sysroot-native/usr/lib/riscv64-lmp-linux/go/src/cmd/go/internal/work/exec.go:281
> +0x2425
> > | cmd/go/internal/work.(*Builder).build(0xc0000030e0, {0xc000000238?,
> 0x0?}, 0xc000172a00)
> > |
>  /srv/oe/build/tmp-lmp/work/x86_64-linux/go-cross-riscv64/1.19-r0/recipe-sysroot-native/usr/lib/riscv64-lmp-linux/go/src/cmd/go/internal/work/exec.go:478
> +0x199
> > | cmd/go/internal/work.(*Builder).Do.func2({0xb2e4f0, 0xc000000238},
> 0xc000172a00)
> > |
>  /srv/oe/build/tmp-lmp/work/x86_64-linux/go-cross-riscv64/1.19-r0/recipe-sysroot-native/usr/lib/riscv64-lmp-linux/go/src/cmd/go/internal/work/exec.go:139
> +0x5fc
> > | cmd/go/internal/work.(*Builder).Do.func3()
> > |
>  /srv/oe/build/tmp-lmp/work/x86_64-linux/go-cross-riscv64/1.19-r0/recipe-sysroot-native/usr/lib/riscv64-lmp-linux/go/src/cmd/go/internal/work/exec.go:201
> +0xb9
> > | created by cmd/go/internal/work.(*Builder).Do
> > |
>  /srv/oe/build/tmp-lmp/work/x86_64-linux/go-cross-riscv64/1.19-r0/recipe-sysroot-native/usr/lib/riscv64-lmp-linux/go/src/cmd/go/internal/work/exec.go:187
> +0x3f5
> > | make: *** [Makefile:244: bin/containerd] Error 2
> > | make: *** Waiting for unfinished jobs....
> >
> > Without this patch everithing look good.
> >
> > Signed-off-by: Jose Quaresma <jose.quaresma@foundries.io>
> > ---
> >  meta/recipes-devtools/go/go-1.19.inc | 1 -
> >  1 file changed, 1 deletion(-)
> >
> > diff --git a/meta/recipes-devtools/go/go-1.19.inc
> b/meta/recipes-devtools/go/go-1.19.inc
> > index f733a807b4..355a82eb9f 100644
> > --- a/meta/recipes-devtools/go/go-1.19.inc
> > +++ b/meta/recipes-devtools/go/go-1.19.inc
> > @@ -5,7 +5,6 @@ FILESEXTRAPATHS:prepend := "${FILE_DIRNAME}/go:"
> >  LIC_FILES_CHKSUM = "file://LICENSE;md5=5d4950ecb7b26d2c5e4e7b4e0dd74707"
> >
> >  SRC_URI += "\
> > -
> file://0001-cmd-go-make-content-based-hash-generation-less-pedan.patch \
> >
> file://0003-allow-GOTOOLDIR-to-be-overridden-in-the-environment.patch \
> >      file://0004-ld-add-soname-to-shareable-objects.patch \
> >
> file://0005-make.bash-override-CC-when-building-dist-and-go_boot.patch \
> > --
> > 2.34.1
> >
> >
> > -=-=-=-=-=-=-=-=-=-=-=-
> > Links: You receive all messages sent to this group.
> > View/Reply Online (#171233):
> https://lists.openembedded.org/g/openembedded-core/message/171233
> > Mute This Topic: https://lists.openembedded.org/mt/94022663/1686489
> > Group Owner: openembedded-core+owner@lists.openembedded.org
> > Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [
> alex.kanavin@gmail.com]
> > -=-=-=-=-=-=-=-=-=-=-=-
> >
>


-- 
Best regards,

José Quaresma

[-- Attachment #2: Type: text/html, Size: 5823 bytes --]

Re: [OE-core] [PATCH 1/3] go: drop patch to make content based hash generation less pedan

[Alexander Kanavin]

On Fri, 30 Sept 2022 at 20:01, Jose Quaresma <quaresma.jose@gmail.com> wrote:
> Forgot to remove the patch! Will send a v2 but I can't run the reproducibility now.

You should, I believe the patch was added in the first place to fix
the reproducibility issues, and won't be accepted if its removal
breaks that.

Alex

Re: [OE-core] [PATCH 1/3] go: drop patch to make content based hash generation less pedan

[Jose Quaresma]

[-- Attachment #1: Type: text/plain, Size: 650 bytes --]

Alexander Kanavin <alex.kanavin@gmail.com> escreveu no dia sexta,
30/09/2022 à(s) 19:02:

> On Fri, 30 Sept 2022 at 20:01, Jose Quaresma <quaresma.jose@gmail.com>
> wrote:
> > Forgot to remove the patch! Will send a v2 but I can't run the
> reproducibility now.
>
> You should, I believe the patch was added in the first place to fix
> the reproducibility issues, and won't be accepted if its removal
> breaks that.
>

Sure, I can do it but only next week.
I tried to understand and come up with a way to fix the issue but don't
understand the real problem behind it.

Jose


> Alex
>


-- 
Best regards,

José Quaresma

[-- Attachment #2: Type: text/html, Size: 1316 bytes --]

Re: [OE-core] [PATCH 1/3] go: drop patch to make content based hash generation less pedan

[Jose Quaresma]

[-- Attachment #1.1: Type: text/plain, Size: 2568 bytes --]

Jose Quaresma via lists.openembedded.org <quaresma.jose=
gmail.com@lists.openembedded.org> escreveu no dia sexta, 30/09/2022 à(s)
19:10:

>
>
> Alexander Kanavin <alex.kanavin@gmail.com> escreveu no dia sexta,
> 30/09/2022 à(s) 19:02:
>
>> On Fri, 30 Sept 2022 at 20:01, Jose Quaresma <quaresma.jose@gmail.com>
>> wrote:
>> > Forgot to remove the patch! Will send a v2 but I can't run the
>> reproducibility now.
>>
>> You should, I believe the patch was added in the first place to fix
>> the reproducibility issues, and won't be accepted if its removal
>> breaks that.
>>
>
> Sure, I can do it but only next week.
> I tried to understand and come up with a way to fix the issue but don't
> understand the real problem behind it.
>

This patch fixes some reproducibility issues as referred by Alex so with
the removal it will come again.
I attach the diffoscope output (diff-html.tar.zst) of the reproducibility
test if anyone is interested.

The affect content is on go-runtime-dev and go-runtime-staticdev:

reproducibleB/tmp/deploy/ipk/core2-64/go-runtime-dev_1.19-r0_core2-64.ipk
./usr/lib/go/pkg/linux_amd64_dynlink/net.a
./usr/lib/go/pkg/linux_amd64_dynlink/os/signal/internal/pty.a
./usr/lib/go/pkg/linux_amd64_dynlink/os/user.a
./usr/lib/go/pkg/linux_amd64_dynlink/plugin.a
./usr/lib/go/pkg/linux_amd64_dynlink/runtime/cgo.a

reproducibleB/tmp/deploy/ipk/core2-64/go-runtime-staticdev_1.19-r0_core2-64.ipk
./usr/lib/go/pkg/linux_amd64/net.a
./usr/lib/go/pkg/linux_amd64/os/signal/internal/pty.a
./usr/lib/go/pkg/linux_amd64/os/user.a
./usr/lib/go/pkg/linux_amd64/plugin.a
./usr/lib/go/pkg/linux_amd64/runtime/cgo.a

So I think we can't remove that patch as it breaks the reproducibility.
Maybe it is better to open a bug because the go is broken in some machines.

I can try to understand better what's going on and try to solve the problem
but it's a little out of my knowledge,
so any help will be much appreciated.

Jose


>
> Jose
>
>
>> Alex
>>
>
>
> --
> Best regards,
>
> José Quaresma
>
> -=-=-=-=-=-=-=-=-=-=-=-
> Links: You receive all messages sent to this group.
> View/Reply Online (#171242):
> https://lists.openembedded.org/g/openembedded-core/message/171242
> Mute This Topic: https://lists.openembedded.org/mt/94022663/5052612
> Group Owner: openembedded-core+owner@lists.openembedded.org
> Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [
> quaresma.jose@gmail.com]
> -=-=-=-=-=-=-=-=-=-=-=-
>
>

-- 
Best regards,

José Quaresma

[-- Attachment #1.2: Type: text/html, Size: 4539 bytes --]

[-- Attachment #2: diff-html.tar.zst --]
[-- Type: application/zstd, Size: 44589 bytes --]