[PATCH core-packages-team] Update Hurd and Mach

  • Done
  • quality assurance status badge
Details
2 participants
  • Janneke Nieuwenhuizen
  • yelninei
Owner
unassigned
Submitted by
yelninei
Severity
normal

Debbugs page

yelninei wrote 10 months ago
(address . guix-patches@gnu.org)(name . Ekaitz Zarraga)(address . ekaitz@elenq.tech)(name . Ludovic Courtès)(address . ludo@gnu.org)(name . Efraim Flashner)(address . efraim@flashner.co.il)(name . Andreas Enge)(address . andreas@enge.fr)(name . Z572)(address . z572@z572.online)(name . Janneke Nieuwenhuizen)(address . janneke@gnu.org)
OPQ6eGy--F-9@tutamail.com
Hello,

This updates hurd and mach for core-packages-team branch to their latest available snapshots. The gnumach update fixes the build of gnumach-headers-boot0 (see https://issues.guix.gnu.org/77709).
I also changed everything to automake@1.17 and the regular texinfo. Should these be separate commits? Also should I separate the commencement.scm part from the hurd.scm part?

Please ignore the prerequisite patch-ids, these are part of WIP fixes for glibc/hurd and other packages to fix more things (again see #77709 for more info).

I hope I did the hashes in commencement.scm correctly, it was a bit tricky to download the sources with the ongoing savannah issues.

I have successfully built these with my WIP glibc patches and "-s i586-gnu" (ignoring some other test failures and adding -Wno-implicit-function-declaration in rumpkernel) and am currently waiting on the cross compiled variants.
Are there additional things that need testing for the 64bit hurd and are there other hurd specific packages that should be updated as well? 

Thanks.
From 7bb361aa6e70f8746745b7a875ac28e3aad3733b Mon Sep 17 00:00:00 2001
Message-ID: <7bb361aa6e70f8746745b7a875ac28e3aad3733b.1746357912.git.yelninei@tutamail.com>
From: Yelninei <yelninei@tutamail.com>
Date: Sun, 4 May 2025 11:25:02 +0000
Subject: [PATCH 1/2] gnu: gnumach: Update to 1.8+git20250304.

* gnu/packages/commencement.scm (gnumach-headers-boot0): Update to 1.8+git20250304.
* gnu/packages/hurd.scm (gnumach-headers): Same.
[native-inputs]: Use automake@1.17.
(gnumach)[native-inputs]: Use automake@1.17.

Change-Id: I74d1a7caf69d2c2d74208eb54c18e3e025e367d7
---
gnu/packages/commencement.scm | 4 ++--
gnu/packages/hurd.scm | 8 ++++----
2 files changed, 6 insertions(+), 6 deletions(-)

Toggle diff (72 lines)
diff --git a/gnu/packages/commencement.scm b/gnu/packages/commencement.scm
index 172a01ffbe..99c3b3833f 100644
--- a/gnu/packages/commencement.scm
+++ b/gnu/packages/commencement.scm
@@ -2733,7 +2733,7 @@ (define gnumach-headers-boot0
(package
(inherit gnumach-headers)
(name "gnumach-headers-boot0")
- (version "1.8+git20240714")
+ (version "1.8+git20250304")
(source
(origin
(inherit (package-source gnumach-headers))
@@ -2747,7 +2747,7 @@ (define gnumach-headers-boot0
"gnumach-" version ".tar.gz"))
(sha256
(base32
- "1bnw5vdbq91zjxklx23qvim40fb0yw1qdxhn9n37jdfypm6q3xir")))))))
+ "1mmjlcbzk1fs64q6vz2p2irp7vgbmyvm5inj3bf522gx50b3v67q")))))))
(native-inputs (list autoconf-boot0 automake-boot0 texinfo-boot0))
(arguments
(substitute-keyword-arguments (package-arguments gnumach-headers)
diff --git a/gnu/packages/hurd.scm b/gnu/packages/hurd.scm
index 2f09a569c2..61ccd19b37 100644
--- a/gnu/packages/hurd.scm
+++ b/gnu/packages/hurd.scm
@@ -57,7 +57,7 @@ (define (hurd-source-url version)
version ".tar.gz"))
(define-public gnumach-headers
- (let ((commit "v1.8+git20240714"))
+ (let ((commit "v1.8+git20250304"))
(package
(name "gnumach-headers")
(version (string-drop commit 1))
@@ -70,7 +70,7 @@ (define-public gnumach-headers
(patches (search-patches "gnumach-version.patch"))
(file-name (git-file-name "gnumach" version))
(sha256
- (base32 "0ykav1kx0bgxcxw04bpcsh5s4531fzdkahjgrlsfs2h3w3vfkga0"))))
+ (base32 "0sq15n7x05qk3rjx5c0dbiijysyvk4gqqqfkc6gjkpsy3mx7a9q0"))))
(build-system gnu-build-system)
(arguments
`(#:phases
@@ -81,7 +81,7 @@ (define-public gnumach-headers
(delete 'build))
#:tests? #f))
(native-inputs
- (list autoconf automake-1.16.5 texinfo))
+ (list autoconf automake texinfo))
(supported-systems %hurd-systems)
(home-page "https://www.gnu.org/software/hurd/microkernel/mach/gnumach.html")
(synopsis "GNU Mach kernel headers")
@@ -270,7 +270,7 @@ (define-public gnumach
(install-file "gnumach.gz" boot))))))))
(native-inputs
(list autoconf
- automake-1.16.5
+ automake
(if (%current-target-system)
(cross-mig (%current-target-system))
mig)

base-commit: e4458ee85aa11e245dd34698f0c83494a9c2b7e7
prerequisite-patch-id: 98adcfd9380022a37fc5add6f1c79a0c8507f91b
prerequisite-patch-id: 9342961af21699c24b27e7d54042f742e5a759d9
prerequisite-patch-id: 26dc56246897b5ab33b3dfc2aea7bdf494570b1e
prerequisite-patch-id: f8749d6a4cacfd491ced0c1ee521ea9b642687dd
prerequisite-patch-id: ac7a97fb34b0459789df7944ded38acc19cd6553
prerequisite-patch-id: 00fc2b4f0602c78f1f352c801c5b358d6686cfc7
--
2.49.0
Attachment: 0002-gnu-hurd-Update-to-0.9.git20250420.patch
yelninei wrote 10 months ago
Re: bug#78241: Acknowledgement ([PATCH core-packages-team] Update Hurd and Mach)
(address . 78241@debbugs.gnu.org)
OPVzYQ9-0B-9@tutamail.com
Cross build hurd and gnumach successfully for i586-pc-gnu and x86_64-pc-gnu from x86_64-linux ontop of a modified core-packages-team branch.
Janneke Nieuwenhuizen wrote 10 months ago
Re: [PATCH core-packages-team] Update Hurd and Mach
(address . yelninei@tutamail.com)(name . Ekaitz Zarraga)(address . ekaitz@elenq.tech)(name . Ludovic Courtès)(address . ludo@gnu.org)(name . Efraim Flashner)(address . efraim@flashner.co.il)(name . Andreas Enge)(address . andreas@enge.fr)(name . Z572)(address . z572@z572.online)(address . guix-patches@gnu.org)
877c2t5187.fsf@gnu.org
Hi,

Toggle quote (5 lines)
> This updates hurd and mach for core-packages-team branch to their
> latest available snapshots. The gnumach update fixes the build of
> gnumach-headers-boot0 (see https://issues.guix.gnu.org/77709). I also
> changed everything to automake@1.17 and the regular texinfo.

Very nice

Toggle quote (3 lines)
> Should these be separate commits? Also should I separate the
> commencement.scm part from the hurd.scm part?

Usually we try to keep patches to commencement separate but for an
identical version bump I think this patch set is OK.

Toggle quote (8 lines)
> I hope I did the hashes in commencement.scm correctly, it was a bit
> tricky to download the sources with the ongoing savannah issues.
>
> I have successfully built these with my WIP glibc patches and "-s
> i586-gnu" (ignoring some other test failures and adding
> -Wno-implicit-function-declaration in rumpkernel) and am currently
> waiting on the cross compiled variants.

Hmm, any idea why adding that flag has become necessary? Should that
maybe be part of this patch set then? If so, you can send a "fixup!" or
"squash!" commit.

Toggle quote (4 lines)
> Are there additional things that need testing for the 64bit hurd and
> are there other hurd specific packages that should be updated as
> well? 

Not that I know of, afaik the 64bit Hurd should "just build".

Toggle quote (14 lines)
> From 7bb361aa6e70f8746745b7a875ac28e3aad3733b Mon Sep 17 00:00:00 2001
> Message-ID: <7bb361aa6e70f8746745b7a875ac28e3aad3733b.1746357912.git.yelninei@tutamail.com>
> From: Yelninei <yelninei@tutamail.com>
> Date: Sun, 4 May 2025 11:25:02 +0000
> Subject: [PATCH 1/2] gnu: gnumach: Update to 1.8+git20250304.

> From 59f3b41dd8e27cb18fe864e3b962eff516fb9c24 Mon Sep 17 00:00:00 2001
> Message-ID: <59f3b41dd8e27cb18fe864e3b962eff516fb9c24.1746357912.git.yelninei@tutamail.com>
> In-Reply-To: <7bb361aa6e70f8746745b7a875ac28e3aad3733b.1746357912.git.yelninei@tutamail.com>
> References: <7bb361aa6e70f8746745b7a875ac28e3aad3733b.1746357912.git.yelninei@tutamail.com>
> From: Yelninei <yelninei@tutamail.com>
> Date: Sun, 4 May 2025 11:25:02 +0000
> Subject: [PATCH 2/2] gnu: hurd: Update to 0.9.git20250420.

Applied to core-packages-team, thanks!

Greetings,
Janneke

--
Janneke Nieuwenhuizen <janneke@gnu.org> | GNU LilyPond https://LilyPond.org
Freelance IT https://www.JoyOfSource.com| Avatar® https://AvatarAcademy.com
yelninei wrote 10 months ago
(name . Janneke Nieuwenhuizen)(address . janneke@gnu.org)(address . 78241@debbugs.gnu.org)
OPaBM5Z--N-9@tutamail.com
Hi Janneke,

May 6, 2025, 13:16 by janneke@gnu.org:

Toggle quote (5 lines)
>
> Hmm, any idea why adding that flag has become necessary? Should that
> maybe be part of this patch set then? If so, you can send a "fixup!" or
> "squash!" commit.
>
The -Wno-implicit-function-declaration is for rumpkernel and unrelated to hurd/mach.
I have just started trying to send some of the patches I have been sitting on for a while.
Here is the full error:
/tmp/guix-build-rumpkernel-0-20211031-0.f1ffd64.drv-0/source/buildrump.sh/src/tools/sed/../../usr.bin/sed/process.c http://buildrump.sh/src/tools/sed/../../usr.bin/sed/process.c: In function 'lputs':
/tmp/guix-build-rumpkernel-0-20211031-0.f1ffd64.drv-0/source/buildrump.sh/src/tools/sed/../../usr.bin/sed/process.c:636:41 http://buildrump.sh/src/tools/sed/../../usr.bin/sed/process.c:636:41: error: implicit declaration of function 'wcwidth' [-Wimplicit-function-declaration]
  636 |                         width = (size_t)wcwidth(wc);


This is caused by using gcc-14 for the HOST_CC and adding gcc-14 as a native-input reproduces the issue on the master branch. Until now the HOST_CC was gcc-11 and only the CROSS_CC has been gcc-14.

In my current checkout I have "fixed" it by adding -V HOST_CFLAGS=-Wno-implicit-function-declaration to the build.sh invocation to silence the warning.
I also saw that rumpkernel has a 20250111 tag in the debian repository but struggled adapting the build system to the new version.
yelninei wrote 10 months ago
(name . Janneke Nieuwenhuizen)(address . janneke@gnu.org)(address . 78241@debbugs.gnu.org)
OPo5Ng7--F-9@tutamail.com
Hello Janneke,

Attached a patch to relax gcc-14 for rumpkernel for core-packages-team. It might be not be needed with the 20250111 tag but I have not found the patience to look into updating it.

It overwrites HOST_CFLAGS which are -O by default. Let me know if you want to add the -Wno-implicit-function-declaration instead of replacing.
May 6, 2025, 15:23 by yelninei@tutamail.com:

Toggle quote (27 lines)
> Hi Janneke,
>
> May 6, 2025, 13:16 by janneke@gnu.org:
>
>>
>> Hmm, any idea why adding that flag has become necessary? Should that
>> maybe be part of this patch set then? If so, you can send a "fixup!" or
>> "squash!" commit.
>>
> The -Wno-implicit-function-declaration is for rumpkernel and unrelated to hurd/mach.
> I have just started trying to send some of the patches I have been sitting on for a while.
>
> Here is the full error:
> /tmp/guix-build-rumpkernel-0-20211031-0.f1ffd64.drv-0/source/> buildrump.sh/src/tools/sed/../../usr.bin/sed/process.c <http://buildrump.sh/src/tools/sed/../../usr.bin/sed/process.c>> : In function 'lputs':
> /tmp/guix-build-rumpkernel-0-20211031-0.f1ffd64.drv-0/source/> buildrump.sh/src/tools/sed/../../usr.bin/sed/process.c:636:41 <http://buildrump.sh/src/tools/sed/../../usr.bin/sed/process.c:636:41>> : error: implicit declaration of function 'wcwidth' [-Wimplicit-function-declaration]
>   636 |                         width = (size_t)wcwidth(wc);
>
>
> This is caused by using gcc-14 for the HOST_CC and adding gcc-14 as a native-input reproduces the issue on the master branch. Until now the HOST_CC was gcc-11 and only the CROSS_CC has been gcc-14.
>
> In my current checkout I have "fixed" it by adding
>  -V HOST_CFLAGS=-Wno-implicit-function-declaration to the build.sh invocation to silence the warning.
>
> I also saw that rumpkernel has a 20250111 tag in the debian repository but struggled adapting the build system to the new version.
>
>
>
From 70d053449042e58a01f7139ae50e487a24a0d133 Mon Sep 17 00:00:00 2001
Message-ID: <70d053449042e58a01f7139ae50e487a24a0d133.1746777814.git.yelninei@tutamail.com>
From: Yelninei <yelninei@tutamail.com>
Date: Fri, 9 May 2025 08:01:19 +0000
Subject: [PATCH] gnu: rumpkernel: Fix build with gcc-14.

* gnu/packages/hurd.scm (rumpkernel): Add -Wno-implicit-function-declaration
for the HOST_CC

Change-Id: I4e86cd9900b56a904190e217c21bcfa279c12a5a
---
gnu/packages/hurd.scm | 1 +
1 file changed, 1 insertion(+)

Toggle diff (22 lines)
diff --git a/gnu/packages/hurd.scm b/gnu/packages/hurd.scm
index 4794c8c376..39bb0629bc 100644
--- a/gnu/packages/hurd.scm
+++ b/gnu/packages/hurd.scm
@@ -850,6 +850,7 @@ (define-public rumpkernel
" -DNO_PCI_MSI_MSIX=yes"
" -DNUSB_DMA=1"
" -DPAE")
+ "-V" "HOST_CFLAGS=-Wno-implicit-function-declaration"
"-V" (string-append
"CWARNFLAGS="
" -Wno-error=maybe-uninitialized"

base-commit: def623300f698f53dceed2bfb01f4dd33bf4f56b
prerequisite-patch-id: 98adcfd9380022a37fc5add6f1c79a0c8507f91b
prerequisite-patch-id: 9342961af21699c24b27e7d54042f742e5a759d9
prerequisite-patch-id: 26dc56246897b5ab33b3dfc2aea7bdf494570b1e
prerequisite-patch-id: f8749d6a4cacfd491ced0c1ee521ea9b642687dd
prerequisite-patch-id: ac7a97fb34b0459789df7944ded38acc19cd6553
prerequisite-patch-id: 00fc2b4f0602c78f1f352c801c5b358d6686cfc7
--
2.49.0
Janneke Nieuwenhuizen wrote 10 months ago
(address . yelninei@tutamail.com)(address . 78241@debbugs.gnu.org)
877c2l22i9.fsf@gnu.org
Hi,

Toggle quote (4 lines)
> Attached a patch to relax gcc-14 for rumpkernel for
> core-packages-team. It might be not be needed with the 20250111 tag
> but I have not found the patience to look into updating it.

That's okay.

Toggle quote (4 lines)
> It overwrites HOST_CFLAGS which are -O by default. Let me know if you
> want to add the -Wno-implicit-function-declaration instead of
> replacing.

Yes, let's keep the explicit "-O" and could you also use

-Wno-error=implicit-function-declaration

that way we'll keep seeing the warning as long as it's there.

Thanks! Greetings,
Janneke

--
Janneke Nieuwenhuizen <janneke@gnu.org> | GNU LilyPond https://LilyPond.org
Freelance IT https://www.JoyOfSource.com| Avatar® https://AvatarAcademy.com
yelninei wrote 10 months ago
(name . Janneke Nieuwenhuizen)(address . janneke@gnu.org)(address . 78241@debbugs.gnu.org)
OQHsXpp--F-9@tutamail.com
Hi,

Fixed in the attached patch. I'll try to get the rest of the core-packages-team fixes ready during the weekend

May 12, 2025, 16:54 by janneke@gnu.org:

Toggle quote (27 lines)
> Hi,
>
>> Attached a patch to relax gcc-14 for rumpkernel for
>> core-packages-team. It might be not be needed with the 20250111 tag
>> but I have not found the patience to look into updating it.
>>
>
> That's okay.
>
>> It overwrites HOST_CFLAGS which are -O by default. Let me know if you
>> want to add the -Wno-implicit-function-declaration instead of
>> replacing.
>>
>
> Yes, let's keep the explicit "-O" and could you also use
>
> -Wno-error=implicit-function-declaration
>
> that way we'll keep seeing the warning as long as it's there.
>
> Thanks! Greetings,
> Janneke
>
> --
> Janneke Nieuwenhuizen <janneke@gnu.org> | GNU LilyPond https://LilyPond.org
> Freelance IT https://www.JoyOfSource.com | Avatar® https://AvatarAcademy.com
>
From 98565fd6e0718d0602b8ba3a0046a9c49cebf19f Mon Sep 17 00:00:00 2001
From: Yelninei <yelninei@tutamail.com>
Date: Fri, 9 May 2025 08:01:19 +0000
Subject: [PATCH] gnu: rumpkernel: Fix build with gcc-14.

* gnu/packages/hurd.scm (rumpkernel): Add -Wno-error=implicit-function-declaration
for the HOST_CC

Change-Id: I4e86cd9900b56a904190e217c21bcfa279c12a5a
---
gnu/packages/hurd.scm | 3 +++
1 file changed, 3 insertions(+)

Toggle diff (16 lines)
diff --git a/gnu/packages/hurd.scm b/gnu/packages/hurd.scm
index 1b2b6553a0..9802469c42 100644
--- a/gnu/packages/hurd.scm
+++ b/gnu/packages/hurd.scm
@@ -851,6 +851,9 @@ (define (noisy-setenv name value)
" -DNO_PCI_MSI_MSIX=yes"
" -DNUSB_DMA=1"
" -DPAE")
+ "-V" (string-append
+ "HOST_CFLAGS=-O"
+ " -Wno-error=implicit-function-declaration")
"-V" (string-append
"CWARNFLAGS="
" -Wno-error=maybe-uninitialized"
--
2.49.0
Janneke Nieuwenhuizen wrote 10 months ago
(address . yelninei@tutamail.com)(address . 78241@debbugs.gnu.org)
87ecwqz3va.fsf@gnu.org
Toggle quote (2 lines)
> Fixed in the attached patch.

Thanks, pushed to core-packages-team.

Toggle quote (3 lines)
> I'll try to get the rest of the
> core-packages-team fixes ready during the weekend

Great; Greeting,
Janneke

--
Janneke Nieuwenhuizen <janneke@gnu.org> | GNU LilyPond https://LilyPond.org
Freelance IT https://www.JoyOfSource.com| Avatar® https://AvatarAcademy.com
yelninei wrote 10 months ago
Closing
(address . 78241-done@debbugs.gnu.org)
OQbP7Jz--F-9@tutamail.com

Closed
?
Your comment

This issue is archived.

To comment on this conversation send an email to 78241@debbugs.gnu.org

To respond to this issue using the mumi CLI, first switch to it
mumi current 78241
Then, you may apply the latest patchset in this issue (with sign off)
mumi am -- -s
Or, compose a reply to this issue
mumi compose
Or, send patches to this issue
mumi send-email *.patch
You may also tag this issue. See list of standard tags. For example, to set the confirmed and easy tags
mumi command -t +confirmed -t +easy
Or, remove the moreinfo tag and set the help tag
mumi command -t -moreinfo -t +help