guix/guix
guix/guix
45
683
Fork 402
Code Issues 756 Pull requests 555 Projects Activity

[hurd-team]: Update mig and gnumach. #6487

Manually merged
janneke merged 18 commits from Yelninei/guix:hurd-team into hurd-team 2026-03-06 19:32:18 +01:00
Conversation 23 Commits 18 Files changed 14 +118 -413
Yelninei commented 2026-02-16 21:31:41 +01:00
Member
Copy link

However it is an abi breaking change (on x86_64 because i586 has no 64bit memebrs in rpcs) and needs new bootstrap files as the old ones dont work with the new format of the mach RPCs..

Ive also included a gnumach update as we are already rebuilding everything.

For commencement.scm I removed the git-fetch-from-tarball as the tarballs are no longer available from savannah. Maybe this needs an immediate SWH fallback when git-download is not available?

We might be able to remove some system-hurd64 code that is now obsolete but this is untested for the reason above.

- The mig update should finally fix guix/guix#417 However it is an abi breaking change (on x86_64 because i586 has no 64bit memebrs in rpcs) and needs new bootstrap files as the old ones dont work with the new format of the mach RPCs.. Ive also included a gnumach update as we are already rebuilding everything. For commencement.scm I removed the `git-fetch-from-tarball` as the tarballs are no longer available from savannah. Maybe this needs an immediate SWH fallback when git-download is not available? We might be able to remove some `system-hurd64` code that is now obsolete but this is untested for the reason above.
Yelninei commented 2026-02-17 17:56:38 +01:00
Author
Member
Copy link

With a rebuild bootstrap files glibc-intermiate fails with a python error? I have never seen this?

python3 -B ../scripts/gen-as-const.py --cc="x86_64-guix-gnu-gcc -std=gnu11 -fgnu89-inline  -g -O2 -Wall -Wwrite-strings -Wundef -Wimplicit-fallthrough -fmerge-all-constants -frounding-math -fno-stack-protector -fno-common -U_FORTIFY_SOURCE -Wno-parentheses -Wstrict-prototypes -Wold-style-definition -fmath-errno      -ftls-model=initial-exec     -I../include -I/tmp/guix-build-glibc-intermediate-2.41.drv-0/build/csu  -I/tmp/guix-build-glibc-intermediate-2.41.drv-0/build  -I../sysdeps/mach/hurd/x86_64  -I../sysdeps/mach/hurd/x86  -I../sysdeps/mach/hurd/x86_64/htl  -I../sysdeps/mach/hurd/htl  -I../sysdeps/hurd/htl  -I../sysdeps/mach/htl  -I../sysdeps/htl/include -I../sysdeps/htl  -I../sysdeps/pthread  -I../sysdeps/x86_64/htl  -I../sysdeps/x86/htl  -I../sysdeps/mach/hurd  -I../sysdeps/gnu  -I../sysdeps/unix/bsd  -I../sysdeps/unix/inet  -I../sysdeps/mach/x86_64  -I../sysdeps/mach/x86  -I../sysdeps/mach/include -I../sysdeps/mach  -I../sysdeps/x86_64/64  -I../sysdeps/x86_64/fpu/multiarch  -I../sysdeps/x86_64/fpu  -I../sysdeps/x86/fpu  -I../sysdeps/x86_64/multiarch  -I../sysdeps/x86_64  -I../sysdeps/x86/include -I../sysdeps/x86  -I../sysdeps/ieee754/float128  -I../sysdeps/ieee754/ldbl-96/include -I../sysdeps/ieee754/ldbl-96  -I../sysdeps/ieee754/dbl-64  -I../sysdeps/ieee754/flt-32  -I../sysdeps/hurd/include -I../sysdeps/hurd  -I../sysdeps/unix  -I../sysdeps/posix  -I../sysdeps/wordsize-64  -I../sysdeps/ieee754  -I../sysdeps/generic -I../hurd -I/tmp/guix-build-glibc-intermediate-2.41.drv-0/build/hurd/ -I../mach -I/tmp/guix-build-glibc-intermediate-2.41.drv-0/build/mach/ -I.. -I../libio -I. -nostdinc -isystem /gnu/store/7gfyy98p9wb09bdn8k2v2d6fx03kkf93-gcc-cross-boot0-14.3.0-lib/lib/gcc/x86_64-guix-gnu/14.3.0/include -isystem /gnu/store/7gfyy98p9wb09bdn8k2v2d6fx03kkf93-gcc-cross-boot0-14.3.0-lib/lib/gcc/x86_64-guix-gnu/14.3.0/include-fixed -isystem /gnu/store/7l3iiabkdxjgw3blz7zpy598q4bgh3lj-hurd-core-headers-0.9.git20251029/include -D_LIBC_REENTRANT -include /tmp/guix-build-glibc-intermediate-2.41.drv-0/build/libc-modules.h -DMODULE_NAME=libc -include ../include/libc-symbols.h       -DTOP_NAMESPACE=glibc -DGEN_AS_CONST_HEADERS \
		   -MD -MP -MF /tmp/guix-build-glibc-intermediate-2.41.drv-0/build/rtld-sizes.h.dT \
		   -MT '/tmp/guix-build-glibc-intermediate-2.41.drv-0/build/rtld-sizes.h.d /tmp/guix-build-glibc-intermediate-2.41.drv-0/build/rtld-sizes.h'" \
	  rtld-sizes.sym > /tmp/guix-build-glibc-intermediate-2.41.drv-0/build/rtld-sizes.hT
Traceback (most recent call last):
  File "../scripts/gen-as-const.py", line 28, in <module>
    import glibcextract
  File "/tmp/guix-build-glibc-intermediate-2.41.drv-0/glibc-2.41/scripts/glibcextract.py", line 23, in <module>
    import subprocess
  File "/gnu/store/qrp7ld67iwp3nma2sn8ff9vf5x7rg0jy-python-boot0-3.5.9/lib/python3.5/subprocess.py", line 129, in <module>
    import selectors
  File "/gnu/store/qrp7ld67iwp3nma2sn8ff9vf5x7rg0jy-python-boot0-3.5.9/lib/python3.5/selectors.py", line 10, in <module>
    import math
ImportError: No module named 'math'
make[2]: *** [../Makerules:271: /tmp/guix-build-glibc-intermediate-2.41.drv-0/build/rtld-sizes.h] Error 1
make[2]: Leaving directory '/tmp/guix-build-glibc-intermediate-2.41.drv-0/glibc-2.41/csu'
make[1]: *** [Makefile:484: csu/subdir_lib] Error 2
make[1]: Leaving directory '/tmp/guix-build-glibc-intermediate-2.41.drv-0/glibc-2.41'
make: *** [Makefile:20: all] Error 2
error: in phase 'build': uncaught exception:
%exception #<&invoke-error program: "make" arguments: ("-j" "1" "--max-load=1") exit-status: 2 term-signal: #f stop-signal: #f> 
phase `build' failed after 123.1 seconds
With a rebuild bootstrap files glibc-intermiate fails with a python error? I have never seen this? ``` python3 -B ../scripts/gen-as-const.py --cc="x86_64-guix-gnu-gcc -std=gnu11 -fgnu89-inline -g -O2 -Wall -Wwrite-strings -Wundef -Wimplicit-fallthrough -fmerge-all-constants -frounding-math -fno-stack-protector -fno-common -U_FORTIFY_SOURCE -Wno-parentheses -Wstrict-prototypes -Wold-style-definition -fmath-errno -ftls-model=initial-exec -I../include -I/tmp/guix-build-glibc-intermediate-2.41.drv-0/build/csu -I/tmp/guix-build-glibc-intermediate-2.41.drv-0/build -I../sysdeps/mach/hurd/x86_64 -I../sysdeps/mach/hurd/x86 -I../sysdeps/mach/hurd/x86_64/htl -I../sysdeps/mach/hurd/htl -I../sysdeps/hurd/htl -I../sysdeps/mach/htl -I../sysdeps/htl/include -I../sysdeps/htl -I../sysdeps/pthread -I../sysdeps/x86_64/htl -I../sysdeps/x86/htl -I../sysdeps/mach/hurd -I../sysdeps/gnu -I../sysdeps/unix/bsd -I../sysdeps/unix/inet -I../sysdeps/mach/x86_64 -I../sysdeps/mach/x86 -I../sysdeps/mach/include -I../sysdeps/mach -I../sysdeps/x86_64/64 -I../sysdeps/x86_64/fpu/multiarch -I../sysdeps/x86_64/fpu -I../sysdeps/x86/fpu -I../sysdeps/x86_64/multiarch -I../sysdeps/x86_64 -I../sysdeps/x86/include -I../sysdeps/x86 -I../sysdeps/ieee754/float128 -I../sysdeps/ieee754/ldbl-96/include -I../sysdeps/ieee754/ldbl-96 -I../sysdeps/ieee754/dbl-64 -I../sysdeps/ieee754/flt-32 -I../sysdeps/hurd/include -I../sysdeps/hurd -I../sysdeps/unix -I../sysdeps/posix -I../sysdeps/wordsize-64 -I../sysdeps/ieee754 -I../sysdeps/generic -I../hurd -I/tmp/guix-build-glibc-intermediate-2.41.drv-0/build/hurd/ -I../mach -I/tmp/guix-build-glibc-intermediate-2.41.drv-0/build/mach/ -I.. -I../libio -I. -nostdinc -isystem /gnu/store/7gfyy98p9wb09bdn8k2v2d6fx03kkf93-gcc-cross-boot0-14.3.0-lib/lib/gcc/x86_64-guix-gnu/14.3.0/include -isystem /gnu/store/7gfyy98p9wb09bdn8k2v2d6fx03kkf93-gcc-cross-boot0-14.3.0-lib/lib/gcc/x86_64-guix-gnu/14.3.0/include-fixed -isystem /gnu/store/7l3iiabkdxjgw3blz7zpy598q4bgh3lj-hurd-core-headers-0.9.git20251029/include -D_LIBC_REENTRANT -include /tmp/guix-build-glibc-intermediate-2.41.drv-0/build/libc-modules.h -DMODULE_NAME=libc -include ../include/libc-symbols.h -DTOP_NAMESPACE=glibc -DGEN_AS_CONST_HEADERS \ -MD -MP -MF /tmp/guix-build-glibc-intermediate-2.41.drv-0/build/rtld-sizes.h.dT \ -MT '/tmp/guix-build-glibc-intermediate-2.41.drv-0/build/rtld-sizes.h.d /tmp/guix-build-glibc-intermediate-2.41.drv-0/build/rtld-sizes.h'" \ rtld-sizes.sym > /tmp/guix-build-glibc-intermediate-2.41.drv-0/build/rtld-sizes.hT Traceback (most recent call last): File "../scripts/gen-as-const.py", line 28, in <module> import glibcextract File "/tmp/guix-build-glibc-intermediate-2.41.drv-0/glibc-2.41/scripts/glibcextract.py", line 23, in <module> import subprocess File "/gnu/store/qrp7ld67iwp3nma2sn8ff9vf5x7rg0jy-python-boot0-3.5.9/lib/python3.5/subprocess.py", line 129, in <module> import selectors File "/gnu/store/qrp7ld67iwp3nma2sn8ff9vf5x7rg0jy-python-boot0-3.5.9/lib/python3.5/selectors.py", line 10, in <module> import math ImportError: No module named 'math' make[2]: *** [../Makerules:271: /tmp/guix-build-glibc-intermediate-2.41.drv-0/build/rtld-sizes.h] Error 1 make[2]: Leaving directory '/tmp/guix-build-glibc-intermediate-2.41.drv-0/glibc-2.41/csu' make[1]: *** [Makefile:484: csu/subdir_lib] Error 2 make[1]: Leaving directory '/tmp/guix-build-glibc-intermediate-2.41.drv-0/glibc-2.41' make: *** [Makefile:20: all] Error 2 error: in phase 'build': uncaught exception: %exception #<&invoke-error program: "make" arguments: ("-j" "1" "--max-load=1") exit-status: 2 term-signal: #f stop-signal: #f> phase `build' failed after 123.1 seconds ```
Yelninei commented 2026-02-17 18:08:17 +01:00
Author
Member
Copy link

math.cpython-35m-i386-gnu.so does not seem to get build

building 'math' extension
gcc -fPIC -Wno-unused-result -Wsign-compare -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -Werror=declaration-after-statement -I./Include -I. -I/tmp/guix-build-python-boot0-3.5.9.drv-0/Python-3.5.9/Include -I/tmp/guix-build-python-boot0-3.5.9.drv-0/Python-3.5.9 -c /tmp/guix-build-python-boot0-3.5.9.drv-0/Python-3.5.9/Modules/mathmodule.c -o build/temp.gnu-0.9-x86_64-3.5/tmp/guix-build-python-boot0-3.5.9.drv-0/Python-3.5.9/Modules/mathmodule.o
/tmp/guix-build-python-boot0-3.5.9.drv-0/Python-3.5.9/Modules/mathmodule.c:70:1: error: static declaration of 'sinpi' follows non-static declaration
70 | sinpi(double x)
| ^~~~~
In file included from /gnu/store/psd26il0iv7lmlgfpbaby5pl9payscjv-glibc-bootstrap-0/include/features.h:524,
from /gnu/store/psd26il0iv7lmlgfpbaby5pl9payscjv-glibc-bootstrap-0/include/bits/libc-header-start.h:33,
from /gnu/store/psd26il0iv7lmlgfpbaby5pl9payscjv-glibc-bootstrap-0/include/limits.h:26,
from ./Include/Python.h:11,
from /tmp/guix-build-python-boot0-3.5.9.drv-0/Python-3.5.9/Modules/mathmodule.c:55:
/gnu/store/psd26il0iv7lmlgfpbaby5pl9payscjv-glibc-bootstrap-0/include/bits/mathcalls.h:81:1: note: previous declaration of 'sinpi' with type 'double(double)'
81 | __MATHCALL_VEC (sinpi,, (Mdouble __x));
| ^~~~~~~~~~~~~~

math.cpython-35m-i386-gnu.so does not seem to get build building 'math' extension gcc -fPIC -Wno-unused-result -Wsign-compare -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -Werror=declaration-after-statement -I./Include -I. -I/tmp/guix-build-python-boot0-3.5.9.drv-0/Python-3.5.9/Include -I/tmp/guix-build-python-boot0-3.5.9.drv-0/Python-3.5.9 -c /tmp/guix-build-python-boot0-3.5.9.drv-0/Python-3.5.9/Modules/mathmodule.c -o build/temp.gnu-0.9-x86_64-3.5/tmp/guix-build-python-boot0-3.5.9.drv-0/Python-3.5.9/Modules/mathmodule.o /tmp/guix-build-python-boot0-3.5.9.drv-0/Python-3.5.9/Modules/mathmodule.c:70:1: error: static declaration of 'sinpi' follows non-static declaration 70 | sinpi(double x) | ^~~~~ In file included from /gnu/store/psd26il0iv7lmlgfpbaby5pl9payscjv-glibc-bootstrap-0/include/features.h:524, from /gnu/store/psd26il0iv7lmlgfpbaby5pl9payscjv-glibc-bootstrap-0/include/bits/libc-header-start.h:33, from /gnu/store/psd26il0iv7lmlgfpbaby5pl9payscjv-glibc-bootstrap-0/include/limits.h:26, from ./Include/Python.h:11, from /tmp/guix-build-python-boot0-3.5.9.drv-0/Python-3.5.9/Modules/mathmodule.c:55: /gnu/store/psd26il0iv7lmlgfpbaby5pl9payscjv-glibc-bootstrap-0/include/bits/mathcalls.h:81:1: note: previous declaration of 'sinpi' with type 'double(double)' 81 | __MATHCALL_VEC (sinpi,, (_Mdouble_ __x)); | ^~~~~~~~~~~~~~
Yelninei commented 2026-02-17 19:03:56 +01:00
Author
Member
Copy link

See this github.com/python/cpython@f57cd8288d in cpython.

A simple s/sinpi/m_sinpi resolves the conflict. Should i add this link somewhere in commencement or in the commit?

See this https://github.com/python/cpython/commit/f57cd8288dbe6aba99c057f37ad6d58f8db75350 in cpython. A simple `s/sinpi/m_sinpi` resolves the conflict. Should i add this link somewhere in commencement or in the commit?
Yelninei commented 2026-02-18 16:13:43 +01:00
Author
Member
Copy link

m4 tests fail to compile with gcc-14, removed the patch that removed the CFLAGS.

m4 tests fail to compile with gcc-14, removed the patch that removed the CFLAGS.
Yelninei commented 2026-02-18 20:31:55 +01:00
Author
Member
Copy link

I added a commit to change crash server to crash-kill as crash-dump-core currently hangs.

Then also added back all the things that I removed before because the errors were casued by exactly this.

this recent commit might at least solve crash-dump-core hanging causing e.g. raise(SIGABRT) to hang.
https://cgit.git.savannah.gnu.org/cgit/hurd/hurd.git/commit/?id=c809367cb503830346bfb1d1fd927c09d3f3cb16

salsa.debian.org/hurd-team/hurd/-@03974e4494

I added a commit to change crash server to crash-kill as crash-dump-core currently hangs. Then also added back all the things that I removed before because the errors were casued by exactly this. this recent commit might at least solve crash-dump-core hanging causing e.g. raise(SIGABRT) to hang. https://cgit.git.savannah.gnu.org/cgit/hurd/hurd.git/commit/?id=c809367cb503830346bfb1d1fd927c09d3f3cb16 https://salsa.debian.org/hurd-team/hurd/-/commit/03974e4494c1b471d31c9c9910dbd56fa4b8c8c0
Yelninei commented 2026-02-19 13:04:12 +01:00
Author
Member
Copy link

Setting crash to crash-kill causes a test failure in guile

Running time.testa
FAIL: time.test: internal-time-units-per-second: versus times and sleep

?

Setting crash to crash-kill causes a test failure in guile Running time.testa FAIL: time.test: internal-time-units-per-second: versus times and sleep ?
Yelninei commented 2026-02-20 09:26:39 +01:00
Author
Member
Copy link

I have built all the things successfully which were causing problems before so this also fixes guix/guix#1041

  • util-linux hanging in configure for ncursesw-config
  • cmake
  • libtool
  • automake
  • git-minimal
  • openssl
  • check
  • tcsh
  • glib

THe things that solve these is the mig update and the patch to make crash-dump-core not hang when handling SIGABRT/abort.

Resolving this also allows simplifying of some tests that were previously skipped because of a hang but now PASS/XFAIL as expected.

I also looked for some system-hurd64? and target-hurd64? checks if they can be removed.

The changes for grep and m4 need the package update on core-packages team to fully get rid of.

There is still some issue with e2fsprogs hanging, this seems fine with -M pc and a lower ram?

The python test hang does not make sense to me right now so I am skipping it for now to move things forward.

The mig change needs updated bootstrap files for x86_64-gnu and old binaries are incompatible. @janneke

I am not sure about removing the git-fetch-from-tarball in commencement?
Savannah does not serve them anymore and we cant determine a hash for SWH fallback of the snapshot taraball.
Is it fine to just consider old daemons as unsupported, should it fallback to SWH immediately if builtin:git-download is not available? @civodul

While building commencement.scm with glibc 2.41 I found an incompatibility with the ancient python-boot0 and libm both providing a sinpi symbol . The fix in python was to just rename theirs so I followed that.
This cahnge is also in the loongarch PR

I have built all the things successfully which were causing problems before so this also fixes guix/guix#1041 - util-linux hanging in configure for ncursesw-config - cmake - libtool - automake - git-minimal - openssl - check - tcsh - glib THe things that solve these is the mig update and the patch to make `crash-dump-core` not hang when handling SIGABRT/abort. Resolving this also allows simplifying of some tests that were previously skipped because of a hang but now PASS/XFAIL as expected. I also looked for some `system-hurd64?` and `target-hurd64?` checks if they can be removed. The changes for grep and m4 need the package update on core-packages team to fully get rid of. There is still some issue with e2fsprogs hanging, this seems fine with `-M pc ` and a lower ram? The python test hang does not make sense to me right now so I am skipping it for now to move things forward. The mig change needs updated bootstrap files for x86_64-gnu and old binaries are incompatible. @janneke I am not sure about removing the `git-fetch-from-tarball` in commencement? Savannah does not serve them anymore and we cant determine a hash for SWH fallback of the snapshot taraball. Is it fine to just consider old daemons as unsupported, should it fallback to SWH immediately if builtin:git-download is not available? @civodul While building commencement.scm with glibc 2.41 I found an incompatibility with the ancient python-boot0 and libm both providing a `sinpi` symbol . The fix in python was to just rename theirs so I followed that. This cahnge is also in the loongarch PR
Yelninei changed title from WIP: [hurd-team]: Update mig and gnumach. to [hurd-team]: Update mig and gnumach. 2026-02-20 09:26:44 +01:00 
* gnu/packages/hurd.scm (gnumach-headers): Update to 1.8+git20260129.

Change-Id: I4007c58d4f99388c78a80eaef783be35208fe16a
Change-Id: I28936a8801ead8aeae76e92df9f31d53bcb5eb73
Yelninei commented 2026-02-23 15:04:34 +01:00
Author
Member
Copy link

Some things that I noticed while rebuilding with the new headers and am not sure what to do about them.

These tests hang but work when using -M pc in qemu, so there maybe is a qemu bug here or somthing else that we are still missing

check is pretty important as it it pretty deep in the dependency graph but I'd rather not put workarounds for packages that work with different qemu settings.

  • check
  • e2fsprogs
Some things that I noticed while rebuilding with the new headers and am not sure what to do about them. These tests hang but work when using `-M pc` in qemu, so there maybe is a qemu bug here or somthing else that we are still missing check is pretty important as it it pretty deep in the dependency graph but I'd rather not put workarounds for packages that work with different qemu settings. - check - e2fsprogs
guix-cuirass-bot approved these changes 2026-02-24 05:04:43 +01:00
Dismissed
guix-cuirass-bot left a comment
Collaborator
Copy link

Results of evaluation 71002 for commit 59d7397f1f:

Results of evaluation [71002](https://pulls.ci.guix.gnu.org/eval/71002) for commit 59d7397f1ff10ef283dda7939f69a16fa66f0447: - Successfully built 2 packages: [`gcc-cross-i586-pc-gnu-toolchain-14.3.0`](https://pulls.ci.guix.gnu.org/build/784702/details), [`gcc-cross-x86_64-pc-gnu-toolchain-14.3.0`](https://pulls.ci.guix.gnu.org/build/784703/details)
guix-cuirass-bot approved these changes 2026-02-25 10:52:56 +01:00
Dismissed
guix-cuirass-bot left a comment
Collaborator
Copy link

Results of evaluation 82710 for commit 9b2ed8d079:

Results of evaluation [82710](https://pulls.ci.guix.gnu.org/eval/82710) for commit 9b2ed8d079c238fd4a1e8cef222cb35c3cbb51e4: - Successfully built 2 packages: [`gcc-cross-i586-pc-gnu-toolchain-14.3.0`](https://pulls.ci.guix.gnu.org/build/825867/details), [`gcc-cross-x86_64-pc-gnu-toolchain-14.3.0`](https://pulls.ci.guix.gnu.org/build/825868/details)
janneke commented 2026-03-01 09:37:50 +01:00
Owner
Copy link

I have rebased this onto latest master, built the bootstrap binaries,
uploaded them to https://lilypond.org/janneke/guix/x86_64-gnu/20260301/,
added a commit updating them for x86_64-gnu, and pushed
it as hurd-team.

Currently I'm building the devel-hurd VM image...but I noticed that upstream
just released/tagged another gnumach this week: v1.8+git20260224.

Do we want to update gnumach again before changing the bootstrap binaries?

Looking at the MIG update, I'm wondering why we're not updating to the
latest commit

8388e79d61c07a0059569851ccdf2e6d9c2761ee
Actually generate code for the 64bit abi when running on a 64bit cpu.
I have rebased this onto latest master, built the bootstrap binaries, uploaded them to https://lilypond.org/janneke/guix/x86_64-gnu/20260301/, added a commit updating them for x86_64-gnu, and pushed it as hurd-team. Currently I'm building the devel-hurd VM image...but I noticed that upstream just released/tagged another gnumach this week: v1.8+git20260224. Do we want to update gnumach again before changing the bootstrap binaries? Looking at the MIG update, I'm wondering why we're not updating to the latest commit 8388e79d61c07a0059569851ccdf2e6d9c2761ee Actually generate code for the 64bit abi when running on a 64bit cpu.
janneke commented 2026-03-01 12:01:28 +01:00
Owner
Copy link
8388e79d61c07a0059569851ccdf2e6d9c2761ee
Actually generate code for the 64bit abi when running on a 64bit cpu.

Oh wait...that's a local commit thit I never upstreamed -- never mind then. Remains the gnumach update; I'll have a look at that. The hurd release that we're using is also pretty old, but there's no new upstream release just yet...

> ``` > 8388e79d61c07a0059569851ccdf2e6d9c2761ee > Actually generate code for the 64bit abi when running on a 64bit cpu. > ``` Oh wait...that's a local commit thit I never upstreamed -- never mind then. Remains the gnumach update; I'll have a look at that. The hurd release that we're using is also pretty old, but there's no new upstream release just yet...
janneke commented 2026-03-02 07:13:41 +01:00
Owner
Copy link

Okay, I upgraded gnumach and hurd, pushed to hurd-team, and created and uploaded new
bootstrap binaries.

Okay, I upgraded gnumach and hurd, pushed to hurd-team, and created and uploaded new bootstrap binaries.
janneke commented 2026-03-03 11:05:06 +01:00
Owner
Copy link

Successfully built a devel-hurd64 image:

./pre-inst-env guix system image --image-type=hurd64-qcow2 --image-size=25G --no-offload gnu/system/examples/devel-hurd64.tmpl

Substitutes available via http://dezyne.org:8181, as always.

I guess updating bootstrap binaries for a new achitecture a couple of times can be OK. I'm just hoping we won't need another update for quite some time, that's also why I updated gnumach and hurd to latest greatest.

WDYT, @Yelninei, @civodul, @efraim?

Successfully built a devel-hurd64 image: ``` ./pre-inst-env guix system image --image-type=hurd64-qcow2 --image-size=25G --no-offload gnu/system/examples/devel-hurd64.tmpl ``` Substitutes available via http://dezyne.org:8181, as always. I guess updating bootstrap binaries for a new achitecture a couple of times can be OK. I'm just hoping we won't need another update for quite some time, that's also why I updated gnumach and hurd to latest greatest. WDYT, @Yelninei, @civodul, @efraim?
Yelninei commented 2026-03-04 19:26:20 +01:00
Author
Member
Copy link

@janneke Does it boot? I tried to cherry pick some ext2fs patches in hopes that the e2fsprogs/check/qemu problem would go away but i ended up with a system that did not boot, but maybe i was missing some

@janneke Does it boot? I tried to cherry pick some ext2fs patches in hopes that the e2fsprogs/check/qemu problem would go away but i ended up with a system that did not boot, but maybe i was missing some
janneke commented 2026-03-05 11:05:54 +01:00
Owner
Copy link

@janneke Does it boot? I tried to cherry pick some ext2fs patches in
hopes that the e2fsprogs/check/qemu problem would go away but i ended
up with a system that did not boot, but maybe i was missing some

As discussed on IRC; yes, it boots!

Before we merge this, I propose to remove

0280d0f8bad Revert "gnu: bootstrap: Update bootstrap binaries for x86_64-gnu, aka the 64-bit Hurd."
ef5f43c94c4 gnu: bootstrap: Update bootstrap binaries for x86_64-gnu, aka the 64-bit Hurd.

and update the commit hash mentioned in commit message

gnu: bootstrap: Update bootstrap binaries for x86_64-gnu, aka the 64-bit Hurd.

On commit:
    4650163903676137b8b42b02f4dc69fb3fe71783
    gnu: hurd: Update to 0.9.git20251029-0.6290b4c.

Then, I guess the build farm's childhurds need to upudated as well.

> @janneke Does it boot? I tried to cherry pick some ext2fs patches in > hopes that the e2fsprogs/check/qemu problem would go away but i ended > up with a system that did not boot, but maybe i was missing some As discussed on IRC; yes, it boots! Before we merge this, I propose to remove ``` 0280d0f8bad Revert "gnu: bootstrap: Update bootstrap binaries for x86_64-gnu, aka the 64-bit Hurd." ef5f43c94c4 gnu: bootstrap: Update bootstrap binaries for x86_64-gnu, aka the 64-bit Hurd. ``` and update the commit hash mentioned in commit message ``` gnu: bootstrap: Update bootstrap binaries for x86_64-gnu, aka the 64-bit Hurd. On commit: 4650163903676137b8b42b02f4dc69fb3fe71783 gnu: hurd: Update to 0.9.git20251029-0.6290b4c. ``` Then, I guess the build farm's childhurds need to upudated as well.
Yelninei commented 2026-03-05 11:36:33 +01:00
Author
Member
Copy link

@janneke I am currently rebuilding commencement.scm for i586 and x86_64.

I hope the (remaining) problems on x86_64 from #6487 (comment) are resolved now.

Those tests are working with "-M PC" in qemu but hang on the -M q35 machine we currently use to get more than 3.5Gib memory.

check is a dependency of python and curl so pretty deep in the graph and the ci childhurds not being able to build them would be problematic.

EDIT:
ext2fs seems to work now, check tests still seems to hang in the check_check test.

@janneke I am currently rebuilding commencement.scm for i586 and x86_64. I hope the (remaining) problems on x86_64 from https://codeberg.org/guix/guix/pulls/6487#issuecomment-10904900 are resolved now. Those tests are working with "-M PC" in qemu but hang on the `-M q35` machine we currently use to get more than 3.5Gib memory. `check` is a dependency of python and curl so pretty deep in the graph and the ci childhurds not being able to build them would be problematic. EDIT: ext2fs seems to work now, check tests still seems to hang in the `check_check` test.
Yelninei changed target branch from master to hurd-team 2026-03-05 18:09:34 +01:00
guix-cuirass-bot requested changes 2026-03-05 18:10:47 +01:00
guix-cuirass-bot left a comment
Collaborator
Copy link

Evaluation failed; see log.

Evaluation [failed](https://pulls.ci.guix.gnu.org/eval/118470); see [log](https://pulls.ci.guix.gnu.org/eval/118470/log/raw).
janneke commented 2026-03-06 08:07:46 +01:00
Owner
Copy link

@guix-cuirass-bot wrote in #6487 (comment):

Evaluation failed; see log.

It's unclear to me which commit you've been looking at, @guix-cuirass-bot? It all seems to work for me?

Anyway, I've cherry-picked the commit, removed the first set of bootstrap binaries and its reversal commit, rebased on latest master and pushed to hurd-team.

@guix-cuirass-bot wrote in https://codeberg.org/guix/guix/pulls/6487#issuecomment-11228509: > Evaluation [failed](https://pulls.ci.guix.gnu.org/eval/118470); see [log](https://pulls.ci.guix.gnu.org/eval/118470/log/raw). It's unclear to me which commit you've been looking at, @guix-cuirass-bot? It all seems to work for me? Anyway, I've cherry-picked the commit, removed the first set of bootstrap binaries and its reversal commit, rebased on latest master and pushed to hurd-team.
Yelninei commented 2026-03-06 08:16:32 +01:00
Author
Member
Copy link

I had a nesting problem error in the initial commit that skipped tests. (i put arguments inside the origin instead of extra field)

ice-9/psyntax.scm:2824:12: In procedure syntax-violation:
Syntax error:
unknown location: %origin: extraneous field initializers (arguments) in form (%origin (method url-fetch) (uri (string-append "https://github.com/libcheck/check/releases/download/" version "/check-" version ".tar.gz")) (hash (content-hash (base32 "02m25y9m46pb6n46s51av62kpd936lkfv3b13kfpckgvmh5lxpm8") sha256)) (patches (list (origin (method url-fetch) (uri (string-append "https://github.com/libcheck/check/commit/" "4fbe702fa4f35bee8a90512f9f59d1441c4ae82e.patch")) (file-name (string-append name "-fix-test-precision-for-ppc.patch")) (sha256 (base32 "04qg1p9afdd6453k18qskazrvscysdcjz9j6w4i6p5x4xyma19v6"))))) (arguments (list #:tests? (and (not (%current-target-system)) (not (system-hurd64?))))))
I had a nesting problem error in the initial commit that skipped tests. (i put arguments inside the origin instead of extra field) ``` ice-9/psyntax.scm:2824:12: In procedure syntax-violation: Syntax error: unknown location: %origin: extraneous field initializers (arguments) in form (%origin (method url-fetch) (uri (string-append "https://github.com/libcheck/check/releases/download/" version "/check-" version ".tar.gz")) (hash (content-hash (base32 "02m25y9m46pb6n46s51av62kpd936lkfv3b13kfpckgvmh5lxpm8") sha256)) (patches (list (origin (method url-fetch) (uri (string-append "https://github.com/libcheck/check/commit/" "4fbe702fa4f35bee8a90512f9f59d1441c4ae82e.patch")) (file-name (string-append name "-fix-test-precision-for-ppc.patch")) (sha256 (base32 "04qg1p9afdd6453k18qskazrvscysdcjz9j6w4i6p5x4xyma19v6"))))) (arguments (list #:tests? (and (not (%current-target-system)) (not (system-hurd64?)))))) ```
Yelninei commented 2026-03-06 09:26:29 +01:00
Author
Member
Copy link

Also it would be great to update the guix package afterwards to have the updated files available from the guix on installed images.

Also it would be great to update the guix package afterwards to have the updated files available from the guix on installed images.
👍 1
Yelninei commented 2026-03-06 18:09:33 +01:00
Author
Member
Copy link

@janneke All packages that previosuly had an error now build

./pre-inst-env guix build  tcsh cmake-minimal python python-minimal openssl libtool automake check util-linux e2fsprogs -e '(@ (gnu packages glib) glib)' git-minimal -s x86_64-gnu --no-grafts -v2 
/gnu/store/f1ql9qkmfb285gvpmgj94c3vqcnnj042-tcsh-6.24.15
/gnu/store/m0sh93pk7365zgi9jlcvkvf96vdgkymj-cmake-minimal-3.31.10
/gnu/store/iaafcd4h5x5k8rzgr50wc3jfhkqfvcwk-python-3.11.14-idle
/gnu/store/aifc33k4vp2x7wl51j0ig9rzwnjwc7ws-python-3.11.14
/gnu/store/ymsr7bxbnacs84p57fqx55y7zj15wd2f-python-3.11.14-tk
/gnu/store/ilkiny0b3sgm2msp7b9i7day41mjrzgj-python-minimal-3.11.14
/gnu/store/scnjy0f4fchhygj38xyjc6ikwbm5v7z1-openssl-3.0.8-doc
/gnu/store/rs8jznr0n72avjg9933dqmn9qz486fzw-openssl-3.0.8
/gnu/store/jdk2pzphgchfsvndsz3nybncpvky12nm-openssl-3.0.8-static
/gnu/store/87j921ls8k44whpzciprxfg68yr3gd4m-libtool-2.4.7
/gnu/store/d40vwz0jf0s28r3l7zv3k6vig14n6jbd-automake-1.17
/gnu/store/b83xpxagg1ahxw8c9dqvmj0yr8s2q31f-check-0.15.2
/gnu/store/snsz51kagclr3vsm80w610gs3z3r18sz-util-linux-2.40.4-lib
/gnu/store/7l1118zcbm1ihjm9v6mda24r29y3b60z-util-linux-2.40.4
/gnu/store/n0awzbviziz677mp4pf9f9l3nk2db3hs-util-linux-2.40.4-static
/gnu/store/hj6sazzkxbrri1fkhfg66s02slzwqajw-e2fsprogs-1.47.2
/gnu/store/zcd9fg70ahldik34bmxbjrh39sazlc7n-glib-2.83.3-bin
/gnu/store/1slxs90nq8vs58k75jghxj4hkjhdn3vs-glib-2.83.3-debug
/gnu/store/cvrxryp8ay26j4rx2fkpisqv6n4xq79w-glib-2.83.3
/gnu/store/60v84y5c5j3afb4k222lgdkx57qj1ar5-glib-2.83.3-static
/gnu/store/7lisihd2mmxv7kfai1hqbkm9i5np71x0-git-minimal-2.52.0
@janneke All packages that previosuly had an error now build ``` ./pre-inst-env guix build tcsh cmake-minimal python python-minimal openssl libtool automake check util-linux e2fsprogs -e '(@ (gnu packages glib) glib)' git-minimal -s x86_64-gnu --no-grafts -v2 /gnu/store/f1ql9qkmfb285gvpmgj94c3vqcnnj042-tcsh-6.24.15 /gnu/store/m0sh93pk7365zgi9jlcvkvf96vdgkymj-cmake-minimal-3.31.10 /gnu/store/iaafcd4h5x5k8rzgr50wc3jfhkqfvcwk-python-3.11.14-idle /gnu/store/aifc33k4vp2x7wl51j0ig9rzwnjwc7ws-python-3.11.14 /gnu/store/ymsr7bxbnacs84p57fqx55y7zj15wd2f-python-3.11.14-tk /gnu/store/ilkiny0b3sgm2msp7b9i7day41mjrzgj-python-minimal-3.11.14 /gnu/store/scnjy0f4fchhygj38xyjc6ikwbm5v7z1-openssl-3.0.8-doc /gnu/store/rs8jznr0n72avjg9933dqmn9qz486fzw-openssl-3.0.8 /gnu/store/jdk2pzphgchfsvndsz3nybncpvky12nm-openssl-3.0.8-static /gnu/store/87j921ls8k44whpzciprxfg68yr3gd4m-libtool-2.4.7 /gnu/store/d40vwz0jf0s28r3l7zv3k6vig14n6jbd-automake-1.17 /gnu/store/b83xpxagg1ahxw8c9dqvmj0yr8s2q31f-check-0.15.2 /gnu/store/snsz51kagclr3vsm80w610gs3z3r18sz-util-linux-2.40.4-lib /gnu/store/7l1118zcbm1ihjm9v6mda24r29y3b60z-util-linux-2.40.4 /gnu/store/n0awzbviziz677mp4pf9f9l3nk2db3hs-util-linux-2.40.4-static /gnu/store/hj6sazzkxbrri1fkhfg66s02slzwqajw-e2fsprogs-1.47.2 /gnu/store/zcd9fg70ahldik34bmxbjrh39sazlc7n-glib-2.83.3-bin /gnu/store/1slxs90nq8vs58k75jghxj4hkjhdn3vs-glib-2.83.3-debug /gnu/store/cvrxryp8ay26j4rx2fkpisqv6n4xq79w-glib-2.83.3 /gnu/store/60v84y5c5j3afb4k222lgdkx57qj1ar5-glib-2.83.3-static /gnu/store/7lisihd2mmxv7kfai1hqbkm9i5np71x0-git-minimal-2.52.0 ```
janneke commented 2026-03-06 19:16:08 +01:00
Owner
Copy link

@Yelninei wrote in #6487 (comment):

@janneke All packages that previosuly had an error now build

./pre-inst-env guix build  tcsh cmake-minimal python python-minimal openssl libtool automake check util-linux e2fsprogs -e '(@ (gnu packages glib) glib)' git-minimal -s x86_64-gnu --no-grafts -v2 
/gnu/store/f1ql9qkmfb285gvpmgj94c3vqcnnj042-tcsh-6.24.15

[..]

That's amazing news! So, do we need anything or can we rebase hurd-team and push to master?

@Yelninei wrote in https://codeberg.org/guix/guix/pulls/6487#issuecomment-11294962: > @janneke All packages that previosuly had an error now build > > ```text > ./pre-inst-env guix build tcsh cmake-minimal python python-minimal openssl libtool automake check util-linux e2fsprogs -e '(@ (gnu packages glib) glib)' git-minimal -s x86_64-gnu --no-grafts -v2 > /gnu/store/f1ql9qkmfb285gvpmgj94c3vqcnnj042-tcsh-6.24.15 [..] That's amazing news! So, do we need anything or can we rebase hurd-team and push to master?
janneke manually merged commit fa3d267dc7 into hurd-team 2026-03-06 19:32:18 +01:00
Sign in to join this conversation.
Reviewers
No reviewers
guix-cuirass-bot
guix/bootstrap
guix/core-packages
guix/hurd
guix/python
Labels
Clear labels   
closure-size

Packages with excesive closure size beyond reasonable

  
deprecation

Changes fall under the deprecation policy.

  
good first issue

Good issue to tackle as a first contribution

  
help-wanted

Contributor is seeking help or advice.

  
kind
bug
Confirmed bug reports.

  
kind
foreign-distro
Issues related to Guix running as package manager on foreign distributions.

  
kind
moreinfo
Need more information for further action.

  
kind
system-test
Failing system tests.

  
kind
wishlist
Feature requests.

  
kind
wont-fix
Out of scope.

  
package
fix

  
patch-upgrade

Only minor patch number increase, x.y.Z, as for semver

  
security-fixes

Security fixes for the package.

  
team-ai

Packages related to AI, ML and LLM.

  
team-audio

Audio related packages.

  
team-beam

Erlang/Elixir/BEAM team.

  
team-bioinformatics

Bioinformatics team.

  
team-bootstrap

Bootstrap

  
team-build-tools

Build tools and parser generators.

  
team-core

Core / Tools / Internals

  
team-core-packages

Core packages: the GNU tool chain, Guile, Coreutils, etc.

  
team-cpp

C and C++ libraries and tools.

  
team-crypto

Secrets, passwords, cryptography and security related packages.

  
team-documentation

Documentation: the manual and cookbook.

  
team-electronics

Electronics and hardware related packages.

  
team-emacs

The extensible, customizable text editor and its ecosystem.

  
team-embedded

Embedded

  
team-games

Packaging programs for amusement.

  
team-gnome

The Gnome desktop environment, along with core technologies such as GLib/GIO, GTK, GStreamer and Webkit.

  
team-go

Go team

  
team-guile

Team for the Guile platform, libraries, and build system.

  
team-hare

Team for the Hare programming language, libraries, and build system.

  
team-haskell

GHC, Hugs, Haskell packages, the "hackage" and "stackage" importers, and the haskell-build-system.

  
team-home

Team for "Guix Home"

  
team-hpc

High performance computing related packages.

  
team-hurd

Team for the Hurd

  
team-installer

Installer script and system installer

  
team-java

The JDK and JRE, the Maven build system, Java packages, the ant-build-system, and the maven-build-system.

  
team-javascript

JavaScript/Node.js packages, the node build system.

  
team-julia

The Julia language, Julia packages, and the julia-build-system.

  
team-kde

The plasma desktop environment, and KDE Applications.

  
team-kernel

Linux-libre kernel team

  
team-lisp

Common Lisp and similar languages, Common Lisp packages and the asdf-build-system.

  
team-localization

Localization of your system to specific languages.

  
team-lua

Lua, Fennel, and related packages.

  
team-lxqt

LXQt desktop environment.

  
team-mate

MATE desktop environment.

  
team-mozilla

Taking care of Icedove and Web Browsers based on Mozilla Thunderbird and Firefox.

  
team-ocaml

The OCaml language, the Dune build system, OCaml packages, the "opam" importer, and the ocaml-build-system.

  
team-perl

Perl team

  
team-python

Python, Python packages, the "pypi" importer, and the python-build-system.

  
team-qa-packages

Quality assurance software (CI/CD, testing, analysis; no direct relation to Guix's QA).

  
team-qt

The Qt toolkit/library and the qt-build-system, as well as some packages using Qt.

  
team-r

The R language, CRAN and Bioconductor repositories, the \"cran\" importer, and the r-build-system.

  
team-racket

The Racket language and Racket-based languages, Racket packages, Racket's variant of Chez Scheme, and development of a Racket build system and importer.

  
team-release

Release process

  
team-reproduciblebuilds

Reproducible Builds tooling and issues that affect any guix packages.

  
team-ruby

Ruby team

  
team-rust

Rust

  
team-science

The main science disciplines and fields related packages (e.g. Astronomy, Chemistry, Math, Physics etc.)

  
team-sugar

Everything related to the Sugar Desktop and learning environment.

  
team-sysadmin

Networking, server clustering, high availability.

  
team-telephony

Telephony packages and services such as Jami, Linphone, etc.

  
team-tex

TeX, LaTeX, XeLaTeX, LuaTeX, TeXLive, the texlive-build-system, and the "texlive" importer.

  
team-translations

Translations

  
team-vcs

Version control packages, services, and Guix download backends.

  
team-xfce

Xfce desktop environment.

  
team-zig

Zig, Zig packages, and the zig-build system

  
user-reviewed

Reviewed and approved.

  
world-rebuild

Affects more than 300 dependent packages

No labels
closure-size
deprecation
good first issue
help-wanted
kind
bug
kind
foreign-distro
kind
moreinfo
kind
system-test
kind
wishlist
kind
wont-fix
package
fix
patch-upgrade
security-fixes
team-ai
team-audio
team-beam
team-bioinformatics
team-bootstrap
team-build-tools
team-core
team-core-packages
team-cpp
team-crypto
team-documentation
team-electronics
team-emacs
team-embedded
team-games
team-gnome
team-go
team-guile
team-hare
team-haskell
team-home
team-hpc
team-hurd
team-installer
team-java
team-javascript
team-julia
team-kde
team-kernel
team-lisp
team-localization
team-lua
team-lxqt
team-mate
team-mozilla
team-ocaml
team-perl
team-python
team-qa-packages
team-qt
team-r
team-racket
team-release
team-reproduciblebuilds
team-ruby
team-rust
team-science
team-sugar
team-sysadmin
team-telephony
team-tex
team-translations
team-vcs
team-xfce
team-zig
user-reviewed
world-rebuild
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
3 participants
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
guix/guix!6487
No description provided.