too long lines were shrinked: clang -pipe -march=native -fno-diagnostics-color -O2 -D_GLIBCXX_ASSERTIONS -Werror=implicit-function-declaration -Werror=implicit-int -fPIC -Icurve/ed25519/nacl_includes -Icurve/ed25519/additions -Icurve/ed25519 -I/usr/include/python3.10 -c curve/ed25519/sc_muladd.c -o /var/tmp/portage/dev-python/py clang -pipe -march=native -fno-diagnostics-color -O2 -D_GLIBCXX_ASSERTIONS -Werror=implicit-function-declaration -Werror=implicit-int -fPIC -Icurve/ed25519/nacl_includes -Icurve/ed25519/additions -Icurve/ed25519 -I/usr/include/python3.10 -c curve/ed25519/sc_reduce.c -o /var/tmp/portage/dev-python/py clang -pipe -march=native -fno-diagnostics-color -O2 -D_GLIBCXX_ASSERTIONS -Werror=implicit-function-declaration -Werror=implicit-int -fPIC -Icurve/ed25519/nacl_includes -Icurve/ed25519/additions -Icurve/ed25519 -I/usr/include/python3.10 -c curve/ed25519/sign.c -o /var/tmp/portage/dev-python/python- clang -pipe -march=native -fno-diagnostics-color -O2 -D_GLIBCXX_ASSERTIONS -Werror=implicit-function-declaration -Werror=implicit-int -fPIC -Icurve/ed25519/nacl_includes -Icurve/ed25519/additions -Icurve/ed25519 -I/usr/include/python3.10 -c curve25519module.c -o /var/tmp/portage/dev-python/python-ax curve25519module.c:161:9: error: incompatible pointer to integer conversion initializing 'Py_ssize_t' (aka 'long') with an expression of type 'void *' [-Wint-conversion] NULL, ^~~~ ------------------------------------------------------------------- This is an unstable amd64 chroot image at a tinderbox (==build bot) name: 17.1_hardened-j4-20220917-070004 ------------------------------------------------------------------- gcc-config -l: [1] x86_64-pc-linux-gnu-12.2.0 * clang/llvm (if any): clang version 15.0.0 Target: x86_64-pc-linux-gnu Thread model: posix InstalledDir: /usr/lib/llvm/15/bin /usr/lib/llvm/15 15.0.0 Python 3.10.7 Available Ruby profiles: (none found) Available Rust versions: [1] rust-bin-1.63.0 * The following VMs are available for generation-2: 1) Eclipse Temurin JDK 11.0.16.1_p1 [openjdk-bin-11] *) Eclipse Temurin JDK 17.0.4.1_p1 [openjdk-bin-17] 3) Eclipse Temurin JDK 8.345_p01 [openjdk-bin-8] Available Java Virtual Machines: [1] openjdk-bin-8 [2] openjdk-bin-11 [3] openjdk-bin-17 system-vm The Glorious Glasgow Haskell Compilation System, version 9.0.2 php cli (if any): GNU Make 4.3 HEAD of ::gentoo commit e8176b4c385fc022aa4f3bd9744cd6ec4632c400 Author: Repository mirror & CI <repomirrorci@gentoo.org> Date: Sun Sep 18 11:16:47 2022 +0000 2022-09-18 11:16:46 UTC emerge -qpvO dev-python/python-axolotl-curve25519 [ebuild N ] dev-python/python-axolotl-curve25519-0.4.1_p2-r2 PYTHON_TARGETS="python3_10 -python3_8 -python3_9"
Created attachment 811228 [details] emerge-info.txt
Created attachment 811231 [details] dev-python:python-axolotl-curve25519-0.4.1_p2-r2:20220918-113418.log
Created attachment 811234 [details] emerge-history.txt
Created attachment 811237 [details] environment
Created attachment 811240 [details] etc.portage.tar.bz2
Created attachment 811243 [details] temp.tar.bz2
Ping.
This will have to wait. I am currently very busy and moving to a new home.
clang-stricter_tinderbox has reproduced this issue with version 0.4.1_p2-r3 - Updating summary.
The bug has been closed via the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=5d93a77c89ebcf55f3673236941668b8f990b337 commit 5d93a77c89ebcf55f3673236941668b8f990b337 Author: Hanno Böck <hanno@gentoo.org> AuthorDate: 2022-12-02 10:00:27 +0000 Commit: Hanno Böck <hanno@gentoo.org> CommitDate: 2022-12-02 10:02:13 +0000 dev-python/python-axolotl-curve25519: Fix compilation with clang. This fixes an invalid type in the python module struct. Also sent upstream. Closes: https://bugs.gentoo.org/871369 Signed-off-by: Hanno Böck <hanno@gentoo.org> .../files/python-axolotl-curve25519-fix-type.patch | 15 +++++++++++++++ .../python-axolotl-curve25519-0.4.1_p2-r3.ebuild | 2 ++ 2 files changed, 17 insertions(+)
This is an example of a build-time fix which needs a new revision: https://github.com/gentoo/devmanual/commit/2b94b64306230dc3595150c5bba87ec0763d93a3.
The bug has been referenced in the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=a41530c6042720c1f05c9836fbfc0446ea533f6d commit a41530c6042720c1f05c9836fbfc0446ea533f6d Author: Conrad Kostecki <conikost@gentoo.org> AuthorDate: 2022-12-04 14:25:33 +0000 Commit: Conrad Kostecki <conikost@gentoo.org> CommitDate: 2022-12-04 14:25:33 +0000 dev-python/python-axolotl-curve25519: force revbump Bug: https://bugs.gentoo.org/871369 Signed-off-by: Conrad Kostecki <conikost@gentoo.org> ...19-0.4.1_p2-r3.ebuild => python-axolotl-curve25519-0.4.1_p2-r4.ebuild} | 0 1 file changed, 0 insertions(+), 0 deletions(-)
Not that it matters a whole lot, but does anyone want to explain why a fix like this needs a revbump? It's a pure compile time issue, 0 and NULL are the same on the binary level.
It's fine for 0/null, it's not fine for others, but it's a lot easier for bits like this to just revbump and find out immediately if there's some problem caused by it. That is, you can often have things which are build-time only fixes but if you slip in the patch, you're not going to find out if there's some problem caused by it (like other people for some crazy reason not being able to build because of it). (But in general, -Wint-conversion does need one, it's just that 0/null is a special case, and it's better to just not overthink it when it's so cheap.)
