Raymond Toy pushed to branch issue-425-correctly-rounded-math-functions-single-float at cmucl / cmucl Commits: f5a811dc by Raymond Toy at 2025-10-17T08:18:53-07:00 Merge branch 'master' into issue-425-correctly-rounded-math-functions - - - - - 07be52cc by Raymond Toy at 2025-10-23T16:30:34-07:00 Fix #449: touch 64-bit-timestamp.txt with desired time - - - - - 443a77eb by Raymond Toy at 2025-10-23T16:30:34-07:00 Merge branch 'issue-449-set-timestamp-on-test-file' into 'master' Fix #449: touch 64-bit-timestamp.txt with desired time Closes #449 See merge request cmucl/cmucl!328 - - - - - c5c80aef by Raymond Toy at 2025-11-06T06:38:25-08:00 21f is not a WIP Remove the heading from the release notes for 21f. Forgot to fix this when doing the 21f release. [skip-ci] - - - - - b764d014 by Raymond Toy at 2025-11-06T07:20:10-08:00 Fix #437: Use realpath to find the executable - - - - - 2a9edb1f by Raymond Toy at 2025-11-06T07:20:11-08:00 Merge branch 'issue-437-find-self' into 'master' Fix #437: Use realpath to find the executable Closes #437 See merge request cmucl/cmucl!329 - - - - - c1904aaa by Raymond Toy at 2025-11-06T08:37:30-08:00 Fix #446: Use C compiler to get errno values and generate UNIX defpackage for errno - - - - - f6df759b by Raymond Toy at 2025-11-06T08:37:30-08:00 Merge branch 'issue-446-use-cc-to-get-errno' into 'master' Fix #446: Use C compiler to get errno values and generate UNIX defpackage for errno Closes #446 See merge request cmucl/cmucl!327 - - - - - b61d6e17 by Raymond Toy at 2025-11-06T17:28:00-08:00 Fix #452: Move bin/errno-foo.lisp to src/tools/ - - - - - 8508d11e by Raymond Toy at 2025-11-06T17:28:00-08:00 Merge branch 'issue-452-move-errno-ref-files' into 'master' Fix #452: Move bin/errno-foo.lisp to src/tools/ Closes #452 See merge request cmucl/cmucl!330 - - - - - ec60199f by Raymond Toy at 2025-11-09T11:30:40-08:00 Address #453: Save analyzer logs even if analyzer fails - - - - - 23ef5ca7 by Raymond Toy at 2025-11-09T11:30:40-08:00 Merge branch 'issue-453-save-analyzer-results' into 'master' Address #453: Save analyzer logs even if analyzer fails See merge request cmucl/cmucl!331 - - - - - 2072ca19 by Raymond Toy at 2025-11-10T05:07:47-08:00 Fix #453: Use correct flags when running the analyzer - - - - - 671d8973 by Raymond Toy at 2025-11-10T05:07:47-08:00 Merge branch 'issue-453-update-make-options-for-analyzer' into 'master' Fix #453: Use correct flags when running the analyzer Closes #453 See merge request cmucl/cmucl!332 - - - - - aca11a4b by Raymond Toy at 2025-11-13T10:56:53-08:00 Fix #373: Add macros to handle temporary files, streams, and directories - - - - - f5484369 by Raymond Toy at 2025-11-13T10:56:53-08:00 Merge branch 'issue-373-handle-temp-files' into 'master' Fix #373: Add macros to handle temporary files, streams, and directories Closes #373 and #375 See merge request cmucl/cmucl!269 - - - - - 420d3ee2 by Raymond Toy at 2025-11-13T13:38:01-08:00 Fix #408: Add C functions to get user name and home dir - - - - - 143a7651 by Raymond Toy at 2025-11-13T13:38:01-08:00 Merge branch 'issue-408-get-user-name-and-homedir' into 'master' Fix #408: Add C functions to get user name and home dir Closes #408, #373, and #375 See merge request cmucl/cmucl!300 - - - - - 7d268755 by Raymond Toy at 2025-11-13T14:21:32-08:00 Update 22a release notes Add recently fixed issues to the notes. [skip-ci] - - - - - b04aaec6 by Raymond Toy at 2025-12-03T18:30:46-08:00 Dummy commit to run pipeline Just removed an extra line at the end of the file. - - - - - 16c334b0 by Raymond Toy at 2025-12-04T15:46:41-08:00 Undo previous change that deleted a blank line Dummy commit to get CI to run to test out concurrent builds. (concurrent=2). - - - - - 25cd44a2 by Raymond Toy at 2025-12-04T17:57:49-08:00 Fix #455: Allow manual pipeline runs - - - - - ae1bdde7 by Raymond Toy at 2025-12-04T17:57:50-08:00 Merge branch 'issue-455-manual-pipeline-run' into 'master' Fix #455: Allow manual pipeline runs Closes #455 See merge request cmucl/cmucl!334 - - - - - fd2e83a1 by Raymond Toy at 2025-12-12T15:38:35-08:00 Fix two important typos in unix-get-username and delete-directory In `unix-get-username`, we were freeing `name` instead of `result`. In `delete-directory`, we misspelled `recursive` as `recusive`. Fix these typos. - - - - - b9494595 by Raymond Toy at 2025-12-12T15:54:05-08:00 Add some tests for unix-get-username and delete-directory Add a simple test that unix-get-username returns something useful. Add a test for delete-directory that tests it deletes directories as specified, include recursively. - - - - - e2b4641a by Raymond Toy at 2025-12-13T14:24:22-08:00 Fix #458: Fix spurious overflow in double-double multiply - - - - - 16097f45 by Raymond Toy at 2025-12-13T14:24:22-08:00 Merge branch 'issue-458-double-double-mult-overflow' into 'master' Fix #458: Fix spurious overflow in double-double multiply Closes #458 See merge request cmucl/cmucl!337 - - - - - f146f87f by Raymond Toy at 2025-12-13T15:11:21-08:00 Add fixed issue #458 to release notes Forgot to do that. [skip-ci] - - - - - 2c36bac8 by Raymond Toy at 2025-12-15T17:34:46-08:00 Use (new) Ubuntu VM to run CI instead of OpenSUSE - - - - - 7ebc2654 by Raymond Toy at 2025-12-15T17:34:46-08:00 Merge branch 'rtoy-add-ubuntu-runner' into 'master' Use (new) Ubuntu VM to run CI instead of OpenSUSE See merge request cmucl/cmucl!339 - - - - - 84435227 by Raymond Toy at 2025-12-15T18:37:49-08:00 Merge branch 'master' into issue-425-correctly-rounded-math-functions - - - - - 6abcc6b4 by Raymond Toy at 2025-12-16T15:57:17-08:00 Use clang to build the core-math routines on Linux Add new variable `CONFIG` to let each build choose what config to use. Currently, we need x86_linux_clang for Linux to be able to build the core-math routines. Eventually, we need to fix it so that gcc can build the core-math routines. - - - - - 6b926281 by Raymond Toy at 2025-12-16T16:38:24-08:00 Try to fix bugs For darwin builds, explicitly set CONFIG to x86_darwin. For Linux cross-compile, we forgot to specify the config with create-target.sh. - - - - - b68a8d74 by Raymond Toy at 2025-12-16T18:40:57-08:00 If the unit tests fail, exit with a non-zero code In `run-tests`, if the results are not empty, exit with a non-zero exit code to tell the caller something didn't pass. - - - - - b939b118 by Raymond Toy at 2025-12-16T18:44:35-08:00 Oops. Use :exitp in print-test-results to exit Revert last change and use the :exitp parameter to print-test-results to exit if there are failures. - - - - - e174f34d by Raymond Toy at 2025-12-16T18:48:08-08:00 Add -u and -p options and exit if lisp-unit tests fail Add -u and -p commandline options to skip the lisp-unit tests or the package-local-nickname tests, respectively. If the lisp-unit tests fail, we want to exit immediately instead of running the package-local-nicknames tests. Without this, it looks like everything is ok if the package-local-nickname tests pass. - - - - - e65f9f5c by Raymond Toy at 2025-12-17T07:04:10-08:00 Fix #457: delete-directory signals errors - - - - - 6034f935 by Raymond Toy at 2025-12-17T07:04:10-08:00 Merge branch 'issue-457-delete-directory-signals-errors' into 'master' Fix #457: delete-directory signals errors Closes #457 See merge request cmucl/cmucl!336 - - - - - 10efd37c by Raymond Toy at 2025-12-17T07:12:47-08:00 Remove extra blank lines Addresses comment from !337: https://gitlab.common-lisp.net/cmucl/cmucl/-/merge_requests/337#note_18529 Don't need to run CI for this. [skip-ci] - - - - - 0996e320 by Raymond Toy at 2025-12-17T13:38:16-08:00 Adjust threshold for log-consistency test Change the threshold for the first test case to 1.42108548d-14 instead of 1.77635684d-15. This worst case value happens for k = 326. - - - - - 53b84f3d by Raymond Toy at 2025-12-17T13:41:47-08:00 Adjust expected value for asinh of most-positive-double-float The expected result has changed by a single bit with core-math. - - - - - 4c8b272e by Raymond Toy at 2025-12-17T13:43:47-08:00 Adjust expected value for acosh of most-positive-double-float The expected result has changed by a single bit with core-math. - - - - - 07fd3bac by Raymond Toy at 2025-12-17T13:45:23-08:00 Let's try gcc with -march=haswell An experiment to see if we can compile the code with gcc if we set -march=hashwell, which is the first processor with the FMA instruction. - - - - - ad844ddd by Raymond Toy at 2025-12-18T06:36:11-08:00 Revert "Let's try gcc with -march=haswell" This reverts commit 07fd3bac2529867a46d9dd610ca4dfaedeb20172. - - - - - 467e0824 by Raymond Toy at 2025-12-18T07:56:25-08:00 Update value for (acosh 4d0) for core-math Core-math returns 2.0634370688955603d0. I'm going to assume that's correct even though (acosh 4w0) is 2.06343706889556054672728117262013w0 And Maxima says: 2.063437068895560546727281172620131871456591449883392499836032693b0 - - - - - b35fd453 by Raymond Toy at 2025-12-18T09:16:17-08:00 Workaround or disable tests failing with core-math For some tests core-math signals a different error. For these, update the expected error condition with the actual error condition. For some tests, core-math doesn't signal an error. Disable these. - - - - - 9c16379a by Raymond Toy at 2025-12-18T11:43:32-08:00 Add -std=gnu11 to CFLAGS for x86_linux config Tested this locally with gcc 15.2.1. The core-math routines can compile along with everything else. - - - - - 885baf0f by Raymond Toy at 2025-12-24T12:43:58-08:00 Merge branch 'master' into issue-425-correctly-rounded-math-functions - - - - - b5fa6d7c by Raymond Toy at 2025-12-26T12:30:10-08:00 Merge branch 'issue-425-correctly-rounded-math-functions' into issue-425-correctly-rounded-math-functions-single-float - - - - - 48 changed files: - .gitignore - .gitlab-ci.yml - bin/build-world.sh - − bin/create-def-unix-error.awk - bin/create-errno.sh - − bin/errno-darwin.lisp - − bin/errno-default.lisp - − bin/errno-linux.lisp - − bin/errno-solaris.lisp - bin/run-ansi-tests.sh - bin/run-unit-tests.sh - src/code/exports.lisp - src/code/extensions.lisp - src/code/filesys.lisp - src/code/os.lisp - src/code/pprint.lisp - src/code/unix.lisp - src/compiler/float-tran-dd.lisp - src/general-info/release-21f.md - src/general-info/release-22a.md - src/i18n/locale/cmucl-unix.pot - src/i18n/locale/cmucl.pot - src/lisp/Config.x86_linux - src/lisp/Darwin-os.c - src/lisp/FreeBSD-os.c - src/lisp/Linux-os.c - src/lisp/NetBSD-os.c - src/lisp/OpenBSD-os.c - src/lisp/lisp.c - src/lisp/os-common.c - src/lisp/solaris-os.c - + src/tools/errno-darwin.lisp - + src/tools/errno-default.lisp - + src/tools/errno-linux.lisp - + src/tools/errno-solaris.lisp - bin/errno-template.lisp → src/tools/errno-template.lisp - + src/tools/unix-pkg-template.lisp - src/tools/worldcom.lisp - src/tools/worldload.lisp - tests/fd-streams.lisp - tests/fdlibm.lisp - tests/os.lisp - tests/pathname.lisp - tests/resources/64-bit-timestamp.txt → tests/resources/64-bit-timestamp-2038.txt - + tests/resources/64-bit-timestamp-2106.txt - tests/run-tests.lisp - tests/trac.lisp - tests/unix.lisp The diff was not included because it is too large. View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/c81e43e930f40b5b267d2db... -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/c81e43e930f40b5b267d2db... You're receiving this email because of your account on gitlab.common-lisp.net.