Differential D43861
OCaml 4.14.1 upgrade. Authored by freebsd_dev.thsi.be on Feb 13 2024, 6:13 AM. Tags None Referenced Files
Details
This revision updates the ports tree to OCaml 4.14.1. Of note: the dune port carries a trivial modification to allow the After this REVISION is accepted, the lang/ocaml-nox11 port should be
Diff Detail
Event TimelineComment Actions Hi, Thanks for your work, this is a much needed update, and not an easy one! At a cursory look the patch looks well composed, but before approval it requires some testing.
While I agree that keeping the commits separated in the revision is useful, especially for you, if the patches require updates, for the final commit to the ports tree I'd be more in favour of one big commit with a succinct description for the individual changes (as in a list of one line descriptions). There are reasons for this. Every commit would require the commit message to be modified to add our standard tags, referencing this revision and all approved by/tested by lines. This is my opinion, maybe other involved committers have different ideas.
Can you explain the logic of this? While I don't really know why it exists, to remove it we need a reason. Comment Actions The idea was that each commit is not introducing any more breakage (tree is buildable at any point), but it is easier to reviewed independently. It doesn't really matter in the end. I suppose that once the review is done, I can push a squashed commits with proper commit message that contains all the information plus the necessary tags.
Well the Graphics module got dropped from 4.13.0. It lives in its own upstream repositoy now. So lang/ocaml has essentially became what lang/ocaml-nox11 was. The later having no more purpose, there is no point in keeping it. The new-old Graphics module (== the new upstream repository) is now packaged in x11-toolkits/ocaml-graphics, and it is much nicer like this (better build system, simplifies lang/ocaml). Comment Actions Hi! I tested your patch in poudriere. In general it is very good from my POV, congrats on your work. I noticed a few minor issues (plist issues mainly) that you should check. I created comments in the Makefiles of the ports having issues.
Comment Actions
Comment Actions Thank you very much for the review!
All those plist issues, along with other newly uncovered and longer standing issues, are now resolved. Comment Actions
Comment Actions Hi, I'm testing the latest patch. I got a new failure in devel/ocaml-findlib: gmake[1]: Leaving directory '/wrkdirs/usr/ports/devel/ocaml-findlib/work/findlib-1.9.6'
cd /wrkdirs/usr/ports/devel/ocaml-findlib/work/findlib-1.9.6/doc && /bin/sh -c '(/usr/bin/find -Ed $1 $3 | /usr/bin/cpio -dumpl $2 >/dev/null 2>&1) && /usr/bin/find -Ed $1 $3 \( -type d -exec /bin/sh -c '\''cd '\''$2'\'' && chmod 755 "$@"'\'' . {} + -o -type f -exec /bin/sh -c '\''cd '\''$2'\'' && chmod 0644 "$@"'\'' . {} + \)' COPYTREE_SHARE \* /wrkdirs/usr/ports/devel/ocaml-findlib/work/stage/usr/local/share/doc/ocaml/findlib "! -regex .*ref-man.*"
/bin/chmod 755 /wrkdirs/usr/ports/devel/ocaml-findlib/work/stage/usr/local/bin/ocamlfind /wrkdirs/usr/ports/devel/ocaml-findlib/work/stage/usr/local/lib/ocaml/site-lib/findlib/make_wizard
chmod: /wrkdirs/usr/ports/devel/ocaml-findlib/work/stage/usr/local/lib/ocaml/site-lib/findlib/make_wizard: No such file or directory
*** Error code 1
Stop.
make: stopped in /usr/ports/devel/ocaml-findlibNot sure what is the cause here. Also, the patch needs rebasing, it fails to apply cleanly on the latest ports tree (due to the changes in graphics/ocaml-images/Makefile)
Comment Actions
Comment Actions Hi, Great work, really! I did another round of testing with the latest patch and got two new failures: in devel/ocaml-logs there are some simple plist issues, some files are missing in the plist: ====> Running Q/A tests (stage-qa) ====> Checking for pkg-plist issues (check-plist) ===> Parsing plist ===> Checking for items in STAGEDIR missing from pkg-plist Error: Orphaned: share/doc/ocaml/logs/CHANGES.md Error: Orphaned: share/doc/ocaml/logs/LICENSE.md Error: Orphaned: share/doc/ocaml/logs/README.md Error: Orphaned: share/doc/ocaml/logs/tags.ml Error: Orphaned: share/doc/ocaml/logs/tool.ml ===> Checking for items in pkg-plist which are not in STAGEDIR ===> Error: Plist issues found. *** Error code 1 Stop. make: stopped in /usr/ports/devel/ocaml-logs =>> Error: check-plist failures detected while games/freetennis is failing with some strange error about a dependency: ===> freetennis-0.4.8_18 depends on file: /usr/local/lib/ocaml/site-lib/lablgtk2/gtk.ml - found ===> Returning to build of freetennis-0.4.8_18 ===> freetennis-0.4.8_18 depends on file: /usr/local/lib/ocaml/lablGL/gl.ml - not found ===> Installing existing package /packages/All/ocaml-lablgl-1.05_6,1.pkg [poudriere.madpilot.net] Installing ocaml-lablgl-1.05_6,1... the most recent version of ocaml-lablgl-1.05_6,1 is already installed ===> freetennis-0.4.8_18 depends on file: /usr/local/lib/ocaml/lablGL/gl.ml - not found *** Error code 1 I suspect after the update the file /usr/local/lib/ocaml/lablGL/gl.ml is not being installed anymore, maybe it has been renamed, so poudriere cannot figure out if the package is already installed. The dependency line should be fixed. BTW the OCAML_DEPENDS variable is only being referenced by BUILD_DEPENDS and AFAIK, it has no other function or purpose, so I think the contents should be directly merged in BUILD_DEPENDS. Comment Actions
Comment Actions Thank you again for the reviews. There had been a bit of fallover from the moving of labltk and lablGL to site-lib/, I should have catched them before... Also Mk/Uses/dune.ml was renamed to Mk/Uses/ocaml.ml, only USES= ocaml:dune is supported now and it still requires to USE_OCAML=yes for bsd.ocaml.mk get pulled in. My ultimate goal is to remove bsd.ocaml.mk, but that's better left for another revision as this ocaml 4.14 update is getting quite massive by now. I've squashed the changes to prepare for the final commit message: only missing the Approved by: and Exp-run by : tags now.
Comment Actions Hi, I'm running with this patch successfully. I also noticed the exp-run at [1] was fine, so I would approve this, patch. I plan to commit the change once all the pending bug reports are approved or timed out. [1] https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=277026 Comment Actions
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||