| 18 Sep 2025 |
 Randy Eckenrode | But currently you should get a separate build because I haven’t fixed that. | 04:20:41 |
 alexfmpe | well the ->gnu64 fails somehow harder, it doesn't even recognize the platform | 04:20:53 |
| alexfmpe |
building for macOS-arm64 but attempting to link with file built for unknown-unsupported file format ( 0x7F 0x45 0x4C 0x46 ...
| 04:21:13 |
| alexfmpe | but, huh, I don't ? | 04:21:35 |
| alexfmpe | ...fjv26-clang-wrapper-19.1.7 in both | 04:21:48 |
| alexfmpe | (still on haskell-updates) | 04:22:03 |
| Randy Eckenrode | Something is really messed up in the Haskell cross environment. | 04:22:04 |
| alexfmpe | story of my life | 04:22:29 |
| alexfmpe | wth | 04:28:32 |
| alexfmpe | $ uname -mo
Darwin arm64
$ nix-build -A haskellPackages.ghc --dry-run
$ nix-build -A pkgsCross.aarch64-darwin.haskellPackages.ghc --dry-run
these 26 derivations will be built:
/nix/store/8xkqxm739x39z14j4jgqakl1wdwx7j1l-apple-sdk-11.3.drv
/nix/store/9d9wxl3994m0dcchml4lj5aw0yvvhr0s-stdenv-darwin.drv
/nix/store/yjak1nd4zqxk6x1s2r5ll8n93z7wraaz-compiler-rt-arm64-apple-darwin-19.1.7.drv
/nix/store/0q7jxpvdaaxzyazmzvxyk5q1xrr2lik2-arm64-apple-darwin-clang-wrapper-19.1.7.drv
/nix/store/hhgmdj05a1x8rw8hrnwzhh7n1vgxcn1g-stdenv-darwin.drv
/nix/store/9268fz8c8qwxml95dk3q5h1drvbyi1xx-libcxx-arm64-apple-darwin-19.1.7.drv
/nix/store/smig5yfrvhz5yi8wfl4lbdlbzrbb2mac-arm64-apple-darwin-clang-wrapper-19.1.7.drv
/nix/store/yg2p67qchszlg259n2vx0dw6ncdjblzn-stdenv-darwin.drv
/nix/store/r4mrx2b3rd00l2957vkmqm6m5w8zcxlp-compiler-rt-libc-arm64-apple-darwin-19.1.7.drv
/nix/store/wsfapxdzldjgby3jqkz1y6afhl527bsh-arm64-apple-darwin-clang-wrapper-19.1.7.drv
/nix/store/qj4d1zzbi1lc14qfn3y0ryxjvaqbjxgm-stdenv-darwin.drv
/nix/store/qpskqafk9mhnl9j1xi4i2094cwj47qs0-apple-sdk-14.4.drv
/nix/store/ys8dagqwfsgyl4ynag8ibhdhd48d6d8r-copyfile-arm64-apple-darwin-213.40.2.drv
/nix/store/0pjkkknmg7jc5vlh9jwbh6riy97mx28f-libutil-arm64-apple-darwin-72.drv
/nix/store/rnzi790dz1cblkhx0kbckrpy24wqsap0-libresolv-arm64-apple-darwin-83.drv
/nix/store/rrgvrg7d2klsm4vimv1h74z223gm53s7-cups-headers-arm64-apple-darwin-2.4.12.drv
/nix/store/w1bd7q5q304m5m5v0nwlh4kvzmvfs6bj-ncurses-arm64-apple-darwin-6.5.drv
/nix/store/jsnfwy9x8zwf6scaw76n12vfjz0cdfap-atf-arm64-apple-darwin-0.23.drv
/nix/store/wzgha4gnd4zlh59l12qb85achyynzqjw-libiconv-arm64-apple-darwin-109.drv
/nix/store/yczpplv5d1gq97xa4z79xlwksk79czya-libsbuf-arm64-apple-darwin-14.1.0.drv
/nix/store/cywcl9fkc9fpp82b9zjxycqj7g3z2zpg-apple-sdk-11.3.drv
/nix/store/mk7fwhz1a6zvmi3fvh60xpg90s2zpx4l-stdenv-darwin.drv
/nix/store/cw5vricxxaf8z18idj20r6m53n6xwf0f-gmp-with-cxx-arm64-apple-darwin-6.3.0.drv
/nix/store/lgk7pmhhxhl0s7i9lhfnl8zy3cbcyy44-hadrian-9.10.3.drv
/nix/store/qc4f1w8v3gs1rddmg6hlvf98r748x5pa-libffi-arm64-apple-darwin-39.drv
/nix/store/h7d5nv0jw25220i0slhgclxk0crzcgqf-arm64-apple-darwin-ghc-9.10.3.drv
| 04:29:13 |
| alexfmpe | this is not what I expected | 04:29:22 |
| alexfmpe | $ nix-build -A hello
/nix/store/ci364cgbwbpww272shfz5mj3y019r9fd-hello-2.12.2
$ nix-build -A pkgsCross.aarch64-darwin.hello
these 23 derivations will be built:
/nix/store/8xkqxm739x39z14j4jgqakl1wdwx7j1l-apple-sdk-11.3.drv
/nix/store/9d9wxl3994m0dcchml4lj5aw0yvvhr0s-stdenv-darwin.drv
| 04:30:02 |
| alexfmpe | ??? | 04:30:03 |
| alexfmpe | shouldn't these be the same | 04:30:07 |
| alexfmpe | same sort of thing happens on release-25.05 | 04:31:10 |
| alexfmpe | top level is cached, pkgsCross.shouldBeMe is something else | 04:31:22 |
| alexfmpe | I could swear there was no difference between hello and pkgsCross.gnu64.hello on my x86_64 nixos | 04:31:53 |
| Randy Eckenrode | Does pkgsCross.x86_64-darwin.haskellPackages.hello build? | 04:34:08 |
| alexfmpe | dunno, can't get through the ghc build | 04:34:30 |
| Randy Eckenrode | Even on 25.05? | 04:34:44 |
| alexfmpe | ah that one I don't know yet, waiting for it | 04:34:53 |
| alexfmpe | next job better buy me a beefy m100 | 04:35:03 |
| alexfmpe | but pkgsCross.aarch64-darwin should be a no-op for me no? | 04:35:59 |
| Randy Eckenrode | Not sure. Something could be set up incorrectly. | 04:36:17 |
| alexfmpe | do...do we expect it to be a no-op ? | 04:36:36 |
| alexfmpe | same ld situation for gnu64 in release-25.05, now building for aarch64-darwin to check that one | 04:52:41 |
| alexfmpe | same thing on release-25.05 | 05:31:34 |
| alexfmpe | I guess it's always been cooked | 05:31:42 |
|  delafthi changed their profile picture. | 14:50:52 |
| delafthi changed their profile picture. | 14:51:15 |
