Thanks for the review, I plan to push Monday.
- Queries
- All Stories
- Search
- Advanced Search
- Transactions
- Transaction Logs
Feed Advanced Search
Advanced Search
Advanced Search
Today
Today
Separate STATICOBJS lines
Yesterday
Yesterday
brooks committed rGf0bd655a615e: libc/quad: further narrow list of i386 symbols (authored by brooks).
libc/quad: further narrow list of i386 symbols
libgcc_s: expose __divmoddi4 on i386
brooks committed rG0b9e3585805e: libgcc_{eh,s}: restore __*_frame_info symbols post llvm18 (authored by brooks).
libgcc_{eh,s}: restore __*_frame_info symbols post llvm18
We could MFC this one, but it's mostly harmless and depends on another commit. Maybe as part of a WITHOUT_UNDEFINED_VERSION MFC batch.
ztest: use ASSERT3P to compare pointers
brooks committed rG6e83504c6031: Revert "lib{c,sys}: normalize export of openat, setcontext, and swapcontext" (authored by brooks).
Revert "lib{c,sys}: normalize export of openat, setcontext, and swapcontext"
I'd lean toward a syscall myself, but don't feel strongly.
Thu, Apr 18
Thu, Apr 18
devel/llvm18: 18.1.4 release
I think I'm going to redo this using readelf. There is some annoying differences between the versions (binutil, elftoolchain, and llvm), but the format feels more manageable and I can see a path to checking for symbols changing type and weak vs strong.
Reading the text that's being removed caused me to submit https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=278441
Seems fine.
brooks committed rG26a09db3add3: Fix incremental build with WITH_NVME newly enabled (authored by brooks).
Fix incremental build with WITH_NVME newly enabled
brooks committed rGee632fb9eb4a: lib{c,sys}: normalize export of openat, setcontext, and swapcontext (authored by brooks).
lib{c,sys}: normalize export of openat, setcontext, and swapcontext
brooks committed rGdf1a09ba524d: libsys: expose a few more symbols for libc's use (authored by brooks).
libsys: expose a few more symbols for libc's use
Wed, Apr 17
Wed, Apr 17
This is slightly too big of a hammer in that rescue.mk will also be regenerated if WITHOUT_NVME is set, but if it's on the way out that seems fine.
devel/llvm-devel: new snapshot
Tue, Apr 16
Tue, Apr 16
libcxxrt: don't export nonexistant symbols
brooks committed rGe0919a4bac2b: libc/arm: export __signalcontext not _signalcontext (authored by brooks).
libc/arm: export __signalcontext not _signalcontext
brooks committed rGb457144ceaf7: libgcc_s: exclude symbols not present on some arches (authored by brooks).
libgcc_s: exclude symbols not present on some arches
brooks committed rGb8df8b0c2e64: libgcc_s: only export 128-bit long double when available (authored by brooks).
libgcc_s: only export 128-bit long double when available
brooks committed rG3f82d8345b17: libgcc_s: only export 128-bit int APIs when available (authored by brooks).
libgcc_s: only export 128-bit int APIs when available
brooks committed rG35b3504ee63e: librpcsec_gss: don't export non-existant symbols (authored by brooks).
librpcsec_gss: don't export non-existant symbols
libc: don't try to export __sF
libc: don't try to export _nsdbtdump
libc: don't try to export mexitcount
brooks committed rG0dd2f4a0cca9: libc: only export hesiod symbols when enabled (authored by brooks).
libc: only export hesiod symbols when enabled
libc: don't try to export mpool_stat
brooks committed rG7dd9070e4425: libc: INTERPOS_SYS macro for interposed syscalls (authored by brooks).
libc: INTERPOS_SYS macro for interposed syscalls
brooks committed rG792081a7300e: lib{c,rt}: use libsys.h for __sys_* declerations (authored by brooks).
lib{c,rt}: use libsys.h for __sys_* declerations
libsys: add a libsys.h
makesyscalls: generate core libsys header
Mon, Apr 15
Mon, Apr 15
sys/acl.h: reduce header pollution
brooks committed rG418b4c2a80a2: acl_to_text: include sys/param.h for MAXLOGNAME (authored by brooks).
acl_to_text: include sys/param.h for MAXLOGNAME
sysproto.h: sys/acl.h -> sys/types.h
brooks committed rGb83d3eb5b798: sys/_{cpu,domain}set.h: don't require sys/param.h (authored by brooks).
sys/_{cpu,domain}set.h: don't require sys/param.h
Reduce reliance on sys/sysproto.h pollution
brooks committed rG101c80fe46b5: sys/malloc.h: move sys/proc.h MALLOC_DECLAREs here (authored by brooks).
sys/malloc.h: move sys/proc.h MALLOC_DECLAREs here
Fri, Apr 12
Fri, Apr 12
Drop sys/malloc.h in kern_time.c (depends on D44766)
- namespace sys/param.h
Wed, Apr 10
Wed, Apr 10
Atomicity is well worth the quite minimal potential cost here.
Thu, Apr 4
Thu, Apr 4
devel/llvm18: Update to 18.1.3
Wed, Apr 3
Wed, Apr 3
devel/llvm-morello: new snapshot
brooks committed R11:2986fba67c34: devel/llvm-devel: handle the edge case where FLANG is on and OPENMP is off (authored by brooks).
devel/llvm-devel: handle the edge case where FLANG is on and OPENMP is off
Tue, Mar 26
Tue, Mar 26
devel/llvm-devel: add missing $
Mon, Mar 25
Mon, Mar 25
Seems reasonable.
devel/llvm-devel: add py-pexpect build dep
Fri, Mar 22
Fri, Mar 22
brooks added reviewers for D44474: acl_to_text: include sys/param.h for MAXLOGNAME: kib, imp, trasz.
Actually squash in sys/queue.h include
- Put ifdef _KERNEL around sys/malloc.h include
- Add sys/types.h and sys/_null.h
- Add a sys/queue.h include after _KERNEL ifdef of sys/malloc.h in sys/acl.h
- Explain sys/malloc.h include in kern_time.c
devel/llvm-devel: update to new snapshot
brooks committed R11:06b0508fd4d5: devel/llvm-{cheri,cheriot,morello}: link llvm-cxx{dump,filt} (authored by brooks).
devel/llvm-{cheri,cheriot,morello}: link llvm-cxx{dump,filt}
devel/llvm18: PLIST_SUB improvements [NFC]
Thu, Mar 21
Thu, Mar 21
riscv: remove more riscv64sf support
msun/riscv: expose fe{disable,enable}except