Bug 283504
| Summary: | www/vimb: Modify the Makefile | ||||||
|---|---|---|---|---|---|---|---|
| Product: | Ports & Packages | Reporter: | mew14930xvi <mew14930xvi> | ||||
| Component: | Individual Port(s) | Assignee: | Zsolt Udvari <uzsolt> | ||||
| Status: | Closed FIXED | ||||||
| Severity: | Affects Only Me | CC: | uzsolt | ||||
| Priority: | --- | ||||||
| Version: | Latest | ||||||
| Hardware: | Any | ||||||
| OS: | Any | ||||||
| URL: | https://portsfallout.com/port/27207/ | ||||||
| See Also: | https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=284585 | ||||||
| Attachments: |
|
||||||
I can build now. Could you please check again? It doesn't always fail. In most cases it will be successful. (In reply to mew14930xvi from comment #2) How can I reproduce it? I don't know. If MAKE_JOBS_UNSAFE=yes, it should not fail. (In reply to mew14930xvi from comment #4) Yes, maybe you've right. As I see the default target is "all" which requires src.subdir-all (and version.h). The src.subdir-all target run "make -C src" and in src/Makefile exists main.o target which requires "../version.h". This src/Makefile doesn't know about Makefile and if the builder machine is "fast" (and parallel) the main.o target reached sooner (in src/Makefile) as version.h target. So I think the "MAKE_JOBS_UNSAFE=yes" would be good solution. Anyway, the version is specified in the port's Makefile, so this is unnecessary processing. It is worth modifying the Makefile even if the build is successful. (In reply to mew14930xvi from comment #6) I agree your modification. I'm thinking: if you generate version.h from Makefile (with or without git) should add MAKE_JOBS_UNSAFE=yes too. What do you think about create a version.h (in files/ with SUB_FILES and SUB_LIST [1]) and place it into WRKSRC (in post-patch target) and remove the version.h-rule from project's Makefile? [1] https://docs.freebsd.org/en/books/porters-handbook/book/#using-sub-files I'll think about it if it fails frequently. (In reply to mew14930xvi from comment #8) Okay. So do you want the actual patch with MAKE_JOBS_UNSAFE=yes? Since I became a maintainer, errors have been reported three times. I'd like to see how it goes before adding MAKE_JOBS_UNSAFE=yes. Okay. Committed, thanks! A commit in branch main references this bug: URL: https://cgit.FreeBSD.org/ports/commit/?id=043f51cb65ad079bc4c551a84689df0488d1a44c commit 043f51cb65ad079bc4c551a84689df0488d1a44c Author: mew14930xvi <mew14930xvi@inbox.lv> AuthorDate: 2024-12-27 18:31:05 +0000 Commit: Zsolt Udvari <uzsolt@FreeBSD.org> CommitDate: 2024-12-27 18:31:05 +0000 www/vimb: Simplify generate version.h Sometimes build fails because of "slow" git-generating of version.h. If it isn't enough should add MAKE_JOBS_UNSAFE=yes. Remove cairo from USE_GNOME. Switch to patch file (instead of post-patch target). PR: 283504 Approved by: submitter is maintainer www/vimb/Makefile | 7 +------ www/vimb/files/patch-Makefile (new) | 15 +++++++++++++++ www/vimb/files/patch-config.mk (new) | 23 +++++++++++++++++++++++ 3 files changed, 39 insertions(+), 6 deletions(-) Thanks. |
Created attachment 256059 [details] modify the Makefile A build error was reported, so modify the Makefile. Simplify USE_GNOME. Switch post-patch: to patch file.