Created attachment 260132 [details] emulators/mame: Update to 0.277 Update to 0.277: * Add FORCE_DRC_C_BACKEND to ${MAKE_ENV}, and retrieve upstream patch to address linking issue with asmjit * Option OpenMP is currently causing ridonkulously long build times during archiving (llvm-ar), so temporarily move it within a BROKEN conditional. Users can still build with OpenMP, they need to add TRYBROKEN to their build environment https://www.mamedev.org/?p=550 poudriere-testport emulators/mame: main-amd64-default: Pass main-i386-default: Pass 13_4-amd64-release: Pass 13_4-i386-release: Pass 13_5-amd64-release: Pass 13_5-i386-release: Pass 14_2-amd64-release: Pass 14_2-i386-release: Pass
Out of curiosity, where did you need long build times with OpenMP? On my box (Ryzen 7900 with -j8 in Poudriere) OpenMP build finished faster than the build without on 14.2-RELEASE amd64.
(In reply to Daniel Engberg from comment #2) Hmm that is great news. I am using a 15-CURRENT host, d14036ea424d5aa3eee20cc6f0b5d7117cc3931b[1] (2025-04-23 04:39:56 +0000). All current releases, and 15-CURRENT jails were taking over hours. The host is a AMD EPYC 7742, with 64 cores available for the build, 124GiB of system memory, TMPFS=all in the poudriere config, and using ccache[3-4]. 1: https://cgit.freebsd.org./src/commit/?id=d14036ea424d5aa3eee20cc6f0b5d7117cc3931b
Ok, interesting. Sounds like a bug on 15-CURRENT :/ Also, is this something we should import? https://gitlab.archlinux.org/archlinux/packaging/packages/mame/-/blob/main/PKGBUILD?ref_type=heads#L63
(In reply to Daniel Engberg from comment #4) > Ok, interesting. Sounds like a bug on 15-CURRENT :/ I will update all the things and see how OpenMP after that. > Also, is this something we should import? > https://gitlab.archlinux.org/archlinux/packaging/packages/mame/-/blob/main/PKGBUILD?ref_type=heads#L63 Sure, will have a look.
Comment on attachment 260132 [details] emulators/mame: Update to 0.277 Will update with a new patch, removing ${BROKEN} from the OPENMP inclusion guard.
(In reply to Daniel Engberg from comment #4) I do not know how Arch Linux uses that upstream commit with 0.277. Their comment[1] references[2] an upstream commit from 2024-09-28, so mame-0.277 should cover this, and browsing the code suggests this too. I will test with ${PATCHFILES} to confirm. 1: https://gitlab.archlinux.org/archlinux/packaging/packages/mame/-/blob/main/PKGBUILD?ref_type=heads#L63 2: https://github.com/mamedev/mame/issues/12825
Created attachment 260827 [details] emulators/mame: Update to 0.277 Update to 0.277: * Add FORCE_DRC_C_BACKEND to ${MAKE_ENV}, to workaround a linking issue with DRC * Retrieve upstream patch to address linking issue with asmjit * Update .ini configuration files, mame, and plugin https://www.mamedev.org/?p=550 PR: 285045 poudriere-testport emulators/mame: 13_4-amd64-release: Pass 13_4-i386-release: Pass 13_5-amd64-release: Pass 13_5-i386-release: Pass 14_2-amd64-release: Pass 14_2-i386-release: Pass main-amd64-default: Pass main-i386-default: Pass
A commit in branch main references this bug: URL: https://cgit.FreeBSD.org/ports/commit/?id=23c010b348e66f91783b94ac658799d080e17d1d commit 23c010b348e66f91783b94ac658799d080e17d1d Author: Alastair Hogge <agh@riseup.net> AuthorDate: 2025-05-31 08:31:45 +0000 Commit: Daniel Engberg <diizzy@FreeBSD.org> CommitDate: 2025-05-31 10:36:02 +0000 emulators/mame: Update to 0.277 * Add FORCE_DRC_C_BACKEND to ${MAKE_ENV}, to workaround a linking issue with DRC * Retrieve upstream patch to address linking issue with asmjit * Update .ini configuration files, mame, and plugin Changelog: https://www.mamedev.org/?p=550 PR: 285045 emulators/mame/Makefile | 9 ++++++--- emulators/mame/distinfo | 8 +++++--- emulators/mame/files/mame.ini.in | 9 ++++++--- emulators/mame/files/plugin.ini.in | 18 +++++++++--------- emulators/mame/pkg-plist | 5 +++-- 5 files changed, 29 insertions(+), 20 deletions(-)
Committed, thanks!
Thank you.