[Bug 294870] graphics/drm-66-kmod, graphics/drm-61-kmod: Fails to build on stable/15 after commit ports d94082fc799f
| Newsgroups | gmane.os.freebsd.devel.x11 |
|---|---|
| Message-ID | <[email protected]/bugzilla/> |
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=294870 Chad Jacob Milios <[email protected]> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |[email protected] --- Comment #10 from Chad Jacob Milios <[email protected]> --- (In reply to Tomoaki AOKI from comment #9) that incidentally will work but i think the underlying issue is that the previous releases were inheriting a system provided list of flags (of which -Wno-address-of-packed-member is one) and adding to it but a logic bug after the addition of compiler_flags.mk precludes that system-provided initialization from happening. upstream also introduced a shape mismatch or variable collision on COMPILER_VERSION which for historical reasons is shaped as 191 by /usr/ports/Mk/Uses/compiler.mk but shaped 190107 by /usr/share/mk/bsd.compiler.mk (the context of submakes of project makefiles which include bsd.kmod.mk -> kmod.mk -> bsd.clang-analyze.mk -> bsd.compiler.mk) and is empty during evaluation by our present local patch-i915_Makefile (which appears to be a no-op as far as i can tell) compiler_flags.mk setting COMPILER_VERSION (e.g. 191) prior to bsd.kmod.mk precludes its setting by bsd.compiler.mk (e.g. 190107) which causes kern.mk to skip the CWARNFLAGS it otherwise would have added after line 15 compiler_flags.mk setting CWARNFLAGS prior to bsd.kmod.mk precludes its initialization at kern.mk line 5 an upstream issue https://github.com/freebsd/drm-kmod/issues/438 has been filed. in the meantime, the attachment of patch-i915_Makefile to Bug 294878 should produce a working build of drm-61-kmod and may work for drm-66-kmod, as is or with minor additions (explicitly opting offending files into -Wno-address-of-packed-member) discussions of where misaligned packed members may be causing performance issues is above my pay grade but i am fairly certain this warning was being ignored for ages and only affects i915 (intel) code, whose users are allergic to performance anyway lol -- You are receiving this mail because: You are the assignee for the bug.