head 1.32; access; symbols; locks; strict; comment @# @; 1.32 date 2026.04.19.02.11.18; author mrg; state Exp; branches; next 1.31; commitid naj8qosku3ZFMvCG; 1.31 date 2026.01.19.02.20.51; author mrg; state Exp; branches; next 1.30; commitid 16pvqiC2yvscJWqG; 1.30 date 2025.12.06.19.39.59; author mrg; state Exp; branches; next 1.29; commitid 8Y8HWUATvJjKTnlG; 1.29 date 2025.11.28.17.02.09; author martin; state Exp; branches; next 1.28; commitid gNl8jT89wuZwhlkG; 1.28 date 2025.11.23.21.27.30; author mrg; state Exp; branches; next 1.27; commitid C8S0byurGU2yUIjG; 1.27 date 2025.11.16.09.01.51; author mrg; state Exp; branches; next 1.26; commitid VPj87UpBhN8n0LiG; 1.26 date 2025.11.16.01.09.58; author mrg; state Exp; branches; next 1.25; commitid 6ZpzmfnVTXOdoIiG; 1.25 date 2025.11.14.05.31.27; author mrg; state Exp; branches; next 1.24; commitid J4YjMSrKJw9rUtiG; 1.24 date 2025.10.29.08.40.57; author mrg; state Exp; branches; next 1.23; commitid eaivUiiplgkmtrgG; 1.23 date 2025.10.07.06.45.32; author mrg; state Exp; branches; next 1.22; commitid rVDK2bjalM3IxBdG; 1.22 date 2025.10.06.06.59.17; author mrg; state Exp; branches; next 1.21; commitid 1wAUxhbrJYskEtdG; 1.21 date 2025.10.06.06.58.05; author mrg; state Exp; branches; next 1.20; commitid paz8pFy6wzWXDtdG; 1.20 date 2025.10.05.05.48.07; author rin; state Exp; branches; next 1.19; commitid nsYtA8TEQiuihldG; 1.19 date 2025.10.05.05.33.23; author rin; state Exp; branches; next 1.18; commitid Ugt4pMSm8T7PbldG; 1.18 date 2025.09.28.02.59.38; author rin; state Exp; branches; next 1.17; commitid RM0JpnVwydsLzqcG; 1.17 date 2025.09.26.18.41.32; author mrg; state Exp; branches; next 1.16; commitid wgrwQO8BI3m6RfcG; 1.16 date 2025.09.26.10.45.21; author rin; state Exp; branches; next 1.15; commitid Qp8T1izjPXPsddcG; 1.15 date 2025.09.24.22.23.50; author mrg; state Exp; branches; next 1.14; commitid sTusS8zGodXm91cG; 1.14 date 2025.09.24.19.40.13; author mrg; state Exp; branches; next 1.13; commitid NsJDf7ONeCIbf0cG; 1.13 date 2025.09.24.19.35.54; author mrg; state Exp; branches; next 1.12; commitid rTBS1tiVmwGwd0cG; 1.12 date 2025.09.24.19.26.41; author skrll; state Exp; branches; next 1.11; commitid CmxZZytkouFHa0cG; 1.11 date 2025.09.24.13.44.38; author rin; state Exp; branches; next 1.10; commitid GzhA9ZhiuYgKgYbG; 1.10 date 2025.09.24.01.41.36; author mrg; state Exp; branches; next 1.9; commitid lBxIuASCZWe8hUbG; 1.9 date 2025.09.23.17.42.56; author skrll; state Exp; branches; next 1.8; commitid eJEmmhKqOkG8DRbG; 1.8 date 2025.09.20.19.01.33; author mrg; state Exp; branches; next 1.7; commitid RfS6Cfg4YeXJ9ubG; 1.7 date 2025.09.20.06.25.15; author mrg; state Exp; branches; next 1.6; commitid lb4V6tGWtIyrXpbG; 1.6 date 2025.09.20.00.34.23; author mrg; state Exp; branches; next 1.5; commitid 4QJUQZFQoHMv1obG; 1.5 date 2025.09.19.22.56.58; author mrg; state Exp; branches; next 1.4; commitid 9SAE6CYqqgiOtnbG; 1.4 date 2025.09.19.09.21.35; author mrg; state Exp; branches; next 1.3; commitid a0VNxkIMNRn4ZibG; 1.3 date 2025.09.19.08.02.26; author mrg; state Exp; branches; next 1.2; commitid iks6fhgfiNVPxibG; 1.2 date 2025.09.19.07.09.34; author mrg; state Exp; branches; next 1.1; commitid XMHErJPYrbEIfibG; 1.1 date 2025.09.17.05.55.30; author mrg; state Exp; branches; next ; commitid xRkkHFTvl2YyT1bG; desc @@ 1.32 log @switch sh3* to GCC 14. the build issue is gone, and testing on landisk/gxemul is identical to the GCC 12 world i built. atf seems to be happy as well. @ text @$NetBSD: README.gcc14,v 1.31 2026/01/19 02:20:51 mrg Exp $ arch/feature list. tools: does build.sh tools work? kernels: does a kernel run? y (yes), b (builds/ready), p (partially), ? (not ready) libgcc: does libgcc build? native-gcc: does a full mknative-gcc complete? sani: has sanitizer been ported to this CPU again for GCC 14? b (build) y (works) make release: does build.sh release complete? (before sanitizer - bs) runs: does the system boot with a full world? atf: does atf run / compare well switched: has port switched? y (yes), n (no/ready), ? (not ready) architecture tools kernels libgcc native-gcc/sani make release runs atf switched ------------ ----- ------- ------ --------------- ------------ ---- --- -------- aarch64 y y y y y y y y y aarch64eb y y y y y y y y y alpha y b y y y y[11] n[14] n[15] ? earmv4 y b y y y y ? ? y earmv4eb y b y y y y ? ? y earmv5 y y y y y y y y y earmv5eb y y y y y y y y y earmv5hf y b y y y y ? ? y earmv5hfeb y b y y y y ? ? y earmv6 y b y y y y ? ? y earmv6eb y b y y y y ? ? y earmv6hf y y y y y y y y y earmv6hfeb y y y y y y y y y earmv7 y b y y y y ? ? y earmv7eb y b y y y y ? ? y earmv7hf y y y y y y y y y earmv7hfeb y y y y y y y y y hppa y b y y y y y y y i386 y b y y y y ? ? y ia64 y y[8] y y y y y ? y m68000 y b y y y n[9] ? ? ? m68k y b y y y y y[2] ? ? mipseb y b y y y y ? ? y mipsel y b y y y y ? ? y mips64eb y y y y y y y y y mips64el y b y y y y ? ? y mipsn64eb y b y y y y ? y y mipsn64el y b y y y y ? ? y powerpc y b y y y y y ? y powerpc64 y b y y y y ? ? y sh3eb y b y y y y ? ? ? sh3el y b y y y y y y y sparc y y y y y y y y y sparc64 y y y y y y y y y vax y y y y y y y y y x86_64 y y y y y y y y y riscv32 y n[7] y y y y ? ? y riscv64 y y y y y y y y y ------------ ----- ------- ------ --------------- ------------ ---- --- architecture tools kernels libgcc native-gcc/sani make release runs atf [2]: GCC isn't stable building pkgsrc nbcompat: gcc -D__DBINTERFACE_PRIVATE -I. -I. -O2 -fno-reorder-blocks -DHAVE_CONFIG_H -c .../libnbcompat-20251029/db/btree/bt_conv.c during GIMPLE pass: slp .../libnbcompat-20251029/db/btree/bt_conv.c: In function '__bt_pgin': .../libnbcompat-20251029/db/btree/bt_conv.c:60:1: internal compiler error: Segmentation fault 60 | __bt_pgin(void *t, pgno_t pg, void *pp) [7]: riscv32 kernels are not really implemented yet, not GCC 12 issue. [8]: GENERIC.SKI boots just as well as GCC 10. [9]: ramdisk wants 4 more KB: Calculated size of `ramdisk.fs.tmp': 614400 bytes, 167 inodes ... 32,nbmakefs: Writing inode 151 (work/./etc/protocols), bytes 0 + 118: No space left on device [11]: alpha needs an -O0 for mopd's file.c to avoid an ICE. [14]: cc1 crashes by jumping into rodata segment when compiling pkgsrc/pkgtools/libnbcompat: In file included from glob.c:12: __glob13.c: In function 'globextend': __glob13.c:800:1: internal compiler error: Illegal instruction 800 | globextend(path, pglob, limit) | ^~~~~~~~~~ [15]: most(?) nfs-related tests timedout. ipf tests fail with incorrect results. some pic_profile variants of tests in usr.bin/c++ crash. CPU vs platform test table (for CPUs with multiple ports). this is "make release" or just kernels. values: y (yes), k (kernels only), n (failed), r (running), ? (not attempted), nx (yes, no x11), nk (no kernels available) CPU platform list --- -------- ---- acorn32 cats epoc32 evbarm-eb evbarm-el hpcarm iyonix netwinder shark zaurus earmv4: k k k y y k k y earm: y y k k earmhf: y y earmv6: y y earmv6hf: y y earmv7: y y earmv7hf: y y amiga atari cesfic hp300 luna68k mac68k mvme68k news68k next68k sun3 virt68k x68k m68k: y k k k k k k k k k k k evbmips emips ews4800mips mipsco newsmips sgimips mipseb: k k k k k k evbmips algor arc cobalt hpcmips pmax mipsel: k k k k k k algor evbmips pmax mips64eb: y mips64el: y y y algor evbmips pmax mipsn64eb: k mipsn64el: k k k amigappc bebox evbppc ibmnws macppc mvmeppc ofppc prep rs6000 sandpoint powerpc: k k k k y k k k k k evbppc macppc ofppc powerpc64: y k n dreamcast evbsh3 hpcsh landisk mmeye sh3eb: k k sh3el: k k k k @ 1.31 log @mips switched to GCC 14 a while back. leaves alpha, sh3, and m68*. (we could probably switch sun2, it does not include gcc itself which is what fails on m68k building pkgsrc nbcompat.) @ text @d1 1 a1 1 $NetBSD: README.gcc14,v 1.30 2025/12/06 19:39:59 mrg Exp $ d48 2 a49 2 sh3eb y b y y y ? ? ? ? sh3el y b y y y n[16] ? ? ? a79 7 [16]: out of bounds checker error external/bsd/ntp/dist/sntp/libopts/libopts.c:24: In function 'canonicalize_pathname', inlined from 'option_pathfind.constprop' at external/bsd/ntp/dist/sntp/libopts/compat/pathfind.c:90:32: external/bsd/ntp/dist/sntp/libopts/compat/pathfind.c:206:17: error: 'strcpy' accessing 1 byte at offsets [-1073741824, 1073741823] and [-1073741824, 1073741823] overlaps 1 byte at offset [-1073741824, 1073741823] [-Werror=restrict] 206 | strcpy( result + i, result + i + 1 ); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @ 1.30 log @note ppc switched. thanks to martin for making it work! @ text @d1 1 a1 1 $NetBSD: README.gcc14,v 1.29 2025/11/28 17:02:09 martin Exp $ d40 6 a45 6 mipseb y b y y y y ? ? ? mipsel y b y y y y ? ? ? mips64eb y y y y y y y ? ? mips64el y b y y y y ? ? ? mipsn64eb y b y y y y ? ? ? mipsn64el y b y y y y ? ? ? @ 1.29 log @Update status for sh3el @ text @d1 1 a1 1 $NetBSD: README.gcc14,v 1.28 2025/11/23 21:27:30 mrg Exp $ d46 2 a47 2 powerpc y b y y y y y ? ? powerpc64 y b y y y y ? ? ? @ 1.28 log @update mips o32 and ppc status. @ text @d1 1 a1 1 $NetBSD: README.gcc14,v 1.27 2025/11/16 09:01:51 mrg Exp $ d49 1 a49 1 sh3el y b y y y ? ? ? ? d80 7 @ 1.27 log @more updates for today. - sh3 builds work now. - pmax n64 kernels build. - update the mips o32 / __atomic_test_and_set info. - note sun2 wants 4KB more for ramdisk.fs @ text @d1 1 a1 1 $NetBSD: README.gcc14,v 1.26 2025/11/16 01:09:58 mrg Exp $ d40 2 a41 2 mipseb y b y y ? ? ? ? ? mipsel y b y y ? n[6] ? ? ? d46 1 a46 1 powerpc y b y y y y n[1] ? ? a58 2 [1]: gcc 14 defines _ARCH_PPC64 with "-mcpu=970 -m32", which gcc 12 did not, and it breaks some assumptions we have about that being in a 64-bit environment only. a64 3 [6]: mips o32 atomic issue, libstdc++'s "test_and_set" calls builtin __atomic_test_and_set() when it doesn't exist libgdb.a(minsyms.o): in function `minimal_symbol_reader::install()': minsyms.c:(.text+0x4224): undefined reference to `__atomic_test_and_set' @ 1.26 log @various updates for gcc 14 support. m68k now builds properly without warnings, but it does not seem to be very stable. pkgsrc building libdb gets an ICE for me in both qemu-virt68k and sun3-tme. add virt68k. fill in some mipsn64 blanks. restore trailing $ for the rcsid. @ text @d1 1 a1 1 $NetBSD$ d48 2 a49 2 sh3eb y b y y n[10] ? ? ? ? sh3el y b y y n[10] ? ? ? ? d67 1 a67 1 [6]: mips o32 also has gdb vs atomic issue: d72 2 a73 1 [9]: Calculated size of `ramdisk.fs.tmp': 614400 bytes, 167 inodes a75 2 [10]: sh3 has missing symbol error in gdb. does does 32-bit mips. minsyms.c:(.text+0x2414): undefined reference to `__atomic_test_and_set' d116 1 a116 1 mipsn64el: k k n @ 1.25 log @arm64 libgcc_s symbol issue is resolved. add mipsn64 (_n_64) variants we support now. @ text @d1 1 a1 1 $NetBSD d39 1 a39 1 m68k y b y y y y[2] y ? ? d61 6 a66 3 [2]: new GCC generated files are too huge, eg: {standard input}: Assembler messages: {standard input}:60465: Warning: .word .L5289-.L5291+00000000 didn't fit d102 2 a103 2 amiga atari cesfic hp300 luna68k mac68k mvme68k news68k next68k sun3 x68k m68k: y k k k k k k k k k k d116 2 a117 2 mipsn64eb: ? mipsn64el: ? ? ? @ 1.24 log @remove old things ia64 builds and switched now. riscv64 was switched a while ago. @ text @d18 2 a19 2 aarch64 y y y y y y y n[13] y aarch64eb y y y y y y y n[13] y d44 2 a74 4 [13]: some pic_profile variants of tests in usr.bin/c++ fails with ld: /usr/lib/libgcc_s.so: undefined reference to `__aarch64_cas8_acq_rel' ld: /usr/lib/libgcc_s.so: undefined reference to `__aarch64_swp8_acq_rel' -- this is odd, these symbols are in libc, and are undefined in libgcc_s.so, but so are eg, abort() and malloc() etc. d112 4 @ 1.23 log @switch to GCC 14 for arm*, arm64*, riscv*, x86, sparc*, hppa. @ text @d37 1 a37 1 ia64 y y[8] y y n[12] ? y ? ? d53 1 a53 1 riscv64 y b y y y y ? ? ? a72 2 [12]: libstdc++ has a weird build issue: ld: /var/obj/ia64/usr/src3/destdir.ia64/usr/lib/libstdc++.so: undefined reference to `__glibcxx_backtrace_simple' a85 42 --- these are GCC 12 issues, but some will remain -- [XX]: on amd64, with very basic test, lsan, ubsan, and asan work, but tsan crashes at start up: ThreadSanitizer: CHECK failed: tsan_rtl.cpp:149 "((reinterpret_cast(this) % 64)) == ((0))" (0x10, 0x0) (tid=21499) ThreadSanitizer:DEADLYSIGNAL ==21499==ERROR: ThreadSanitizer: SEGV on unknown address 0x000000000017 (pc 0x7f7ff4e5e14e bp 0x7f7fffffdbd0 sp 0x7f7fffffdb40 T21499) ==21499==The signal is caused by a READ memory access. ==21499==Hint: address points to the zero page. ThreadSanitizer:DEADLYSIGNAL ThreadSanitizer: nested bug in the same thread, aborting. [XX]: "dd count=1", "env LC_CTYPE=en_US.UTF-8 locale", and "tests/crypto/libcrypto/h_evp_test evpmd_blake.txt" abort. adding -fno-stack-protector to strsuftoll.c, citrus_module.c, and blake2_prov.c, respectively, works around the problem. -- workarounds commited. [XX]: sanitizers crash early: Program received signal SIGSEGV, Segmentation fault. (gdb) bt #0 0x000000004064265c in __sanitizer::CheckASLR () at /usr/src/external/gpl3/gcc/dist/libsanitizer/sanitizer_common/sanitizer_linux.cpp:2181 #1 0xffffffffffffc704 in ?? () Backtrace stopped: previous frame identical to this frame (corrupt stack?) -- some wip patches make this less bad [XX]: (NOT A REGRESSION) libtsan is newly introduced, which fails with: FATAL: ThreadSanitizer: unexpected memory mapping 0x200100000-0x200101000 [XX]: sanitizers fail to build: sanitizer_flag_parser.h:141:72: error: format '%zx' expects argument of type 'size_t', but argument 4 has type 'unsigned int' [-Werror=format=] [XX]: (NOT A REGRESSION) c++ exception handling is broken. mechanically applying http://www.execsw.org/netbsd/changeset.cgi?id=20140318T234341Z.c59ac4642e7d4b345f6dc5aa9887e9e1f9aa459a#src/gnu/dist/gcc4/gcc/config/vax/vax.c does not help (with adjustments for gcc 10.5.0, which is broken, too): (gdb) invalid-command [1] Illegal instruction (core dumped) gdb [XX]: (maybe not related to [24]) native gcc is broken: simh# cc hello.c during RTL pass: final hello.c: In function 'main': hello.c:9:1: internal compiler error: Segmentation fault [XX]: divide by zero with ubsan hits CheckLocked assertion failure, and stalls at LockSlow indefinitely: (gdb) bt #0 __sanitizer::StaticSpinMutex::LockSlow at sanitizer_mutex.cpp:24 ... #17 __ubsan::ScopedReport::CheckLocked ubsan_diag.h:255 #18 __ubsan::Diag::~Diag ubsan_diag.cpp:347 #19 handleDivremOverflowImpl ubsan_diag.h:210 #20 __ubsan::__ubsan_handle_divrem_overflow ubsan_handlers.cpp:320 @ 1.22 log @add a note about the arm64 vs __aarch64_{cas,swp}8_acq_rel issue. @ text @d18 2 a19 2 aarch64 y y y y y y y n[13] ? aarch64eb y y y y y y y n[13] ? d21 16 a36 16 earmv4 y b y y y y ? ? ? earmv4eb y b y y y y ? ? ? earmv5 y y y y y y y y ? earmv5eb y y y y y y y y ? earmv5hf y b y y y y ? ? ? earmv5hfeb y b y y y y ? ? ? earmv6 y b y y y y ? ? ? earmv6eb y b y y y y ? ? ? earmv6hf y y y y y y y y ? earmv6hfeb y y y y y y y y ? earmv7 y b y y y y ? ? ? earmv7eb y b y y y y ? ? ? earmv7hf y y y y y y y y ? earmv7hfeb y y y y y y y y ? hppa y b y y y y y y ? i386 y b y y y y ? ? ? d48 5 a52 5 sparc y y y y y y y y ? sparc64 y y y y y y y y ? vax y y y y y y y y ? x86_64 y y y y y y y y ? riscv32 y n[7] y y y y ? ? ? @ 1.21 log @note that 32 bit mips kernels build. @ text @d78 1 @ 1.20 log @README.gcc14: Update terrible status of alpha - Unable to build pkgsrc/pkgtools/nbcompat. - Most(?) nfs-related tests timedout. - ipf tests fail with incorrect results. - Some pic_profile variants tests in usr.bin/c++ crash. @ text @d40 2 a41 2 mipseb y ? y y ? ? ? ? ? mipsel y ? y y ? n[6] ? ? ? @ 1.19 log @README.gcc14: Update aarch64{,eb} Now, aarch64eb works at a level of little-endian counterpart. Full ATF completes succesfully, except for some failures for pic_profile variants of usr.bin/c++: ld: /usr/lib/libgcc_s.so: undefined reference to `__aarch64_cas8_acq_rel' ld: /usr/lib/libgcc_s.so: undefined reference to `__aarch64_swp8_acq_rel' @ text @d20 1 a20 1 alpha y b y y y y[11] ? ? ? d78 8 a85 1 @ 1.18 log @README.gcc14: Update 32-bit arm platforms evbarmv5{,eb} and evbarmv{6,7}hf{,eb} just work fine, and there's no new regression for full ATF. @ text @d18 2 a19 2 aarch64 y y y y y y y ? ? aarch64eb y y y y y y n[5] ? ? a61 6 [5]: after extracting base, 'tar + gzip' combo sometimes crashes. kernel works. eg: pax -rzpe -f /ns3/base.tgz pax -rzpe -f /ns3/comp.tgz gzip: error writing to output: Broken pipe [1] Abort trap (core dumped) ${doit} pax -rzpe ${verbose} -f "${setfile}" most things seem to work when used in a chroot. for above it is pax (or tar) that crashes. d75 3 @ 1.17 log @update the macppc / _ARCH_PPC64 entry: was missing a flag @ text @d23 2 a24 2 earmv5 y b y y y y ? ? ? earmv5eb y b y y y y ? ? ? d29 2 a30 2 earmv6hf y b y y y y ? ? ? earmv6hfeb y b y y y y ? ? ? d33 2 a34 2 earmv7hf y b y y y y ? ? ? earmv7hfeb y b y y y y ? ? ? @ 1.16 log @README.gcc14: Update vax -O0 hacks are gone. No new regressions for ATF. Thanks kalvisd@@ :) @ text @d57 1 a57 1 [1]: gcc 14 defines _ARCH_PPC64 with -mcpu=970, which gcc 12 did not, and it breaks some @ 1.15 log @evbmips64-eb boots and runs (and while yet updated, atf seems to be ok so far.) @ text @d50 1 a50 1 vax y y[4] y y y y y ? ? a61 2 [4]: my kernel has 2 files built with -O0 to avoid compiler ICE issues: vsbus.c and ncr53c9x.c. gallium and a couple of other parts of userland trigger same problem. @ 1.14 log @note that powerpc has a problem: _ARCH_PPC64 defined when not in gcc 12. @ text @d42 1 a42 1 mips64eb y b y y y y ? ? ? @ 1.13 log @update arm64 status: little endian seems fine, will be atf'ing soon. big endian still fails to run gzip and tar or pax with the latter crashing. @ text @d44 1 a44 1 powerpc y b y y y y ? ? ? d57 2 @ 1.12 log @Update hppa status @ text @d18 1 a18 1 aarch64 y ? y y ? n[1] ? ? ? a56 1 [1]: there is a ctfconvert vs arm problem in libgcc, and probably more d62 6 a67 1 [5]: after extracting base, 'tar + gzip' combo sometimes crashes. kernel works. @ 1.11 log @README.gcc14: Update m68k amiga (including boot loaders and X server) works just fine for me. Now, it runs full ATF. @ text @d35 1 a35 1 hppa y b y y y y ? ? ? @ 1.10 log @status update. - arm64eb sort of works, but weird issue with gzip + tar|pax - alpha, hppa builds - ia64 is slightly less broken - sh3 & mips have missing __atomic_test_and_set @ text @d39 1 a39 1 m68k y b y y y n[2] ? ? ? d137 1 a137 1 m68k: k k k k k k k k k k k @ 1.9 log @Update hppa status @ text @d19 2 a20 2 aarch64eb y y y y y y ? ? ? alpha y b y y n[2] ? ? ? ? d35 1 a35 1 hppa y b y y ? y ? ? ? d37 1 a37 1 ia64 y y[8] y y n[2] ? y ? ? d39 1 a39 1 m68k y b y y n[2] ? ? ? ? d41 1 a41 1 mipsel y ? y y ? n[5,6] ? ? ? d46 2 a47 2 sh3eb y b y y n[2] ? ? ? ? sh3el y b y y n[2] ? ? ? ? d50 2 a51 2 vax y y[4] y y n[2] ? ? ? ? x86_64 y y y y y ? y y ? d58 6 a63 14 [2]: sanitizer_asm.h needs significant porting: /usr/src3/external/gpl3/gcc/dist/libsanitizer/interception/interception.h:208:8: error: expected ')' before 'C_ASM_TAIL_CALL' 208 | C_ASM_TAIL_CALL(SANITIZER_STRINGIFY(TRAMPOLINE(func)), \ | ^~~~~~~~~~~~~~~ /usr/src3/external/gpl3/gcc/dist/libsanitizer/interception/interception.h:305:3: note: in expansion of macro 'DECLARE_WRAPPER' 305 | DECLARE_WRAPPER(ret_type, func, __VA_ARGS__) \ | ^~~~~~~~~~~~~~~ /usr/src3/external/gpl3/gcc/dist/libsanitizer/sanitizer_common/sanitizer_common_interceptors.inc:380:1: note: in expansion of macro 'INTERCEPTOR' 380 | INTERCEPTOR(SIZE_T, strlen, const char *s) { | ^~~~~~~~~~~ [4]: my kernel has 3 or 4 files built with -O0 to avoid compiler ICE issues [5]: build fails in libstdc++: libstdc++-v3/include/bits/utility.h:234:31: error: missing binary operator before token "(" 234 | #if _GLIBCXX_USE_BUILTIN_TRAIT(__type_pack_element) d72 5 @ 1.8 log @status update - arm64eb boots = sun2 almost builds (ramdisk full error) - sparc64 and amd64 atf work - all the 32-bit arms build @ text @d35 1 a35 1 hppa y ? y y ? n[3] ? ? ? a67 3 [3]: looks like some codegen issue in libgcc, eg: {standard input}:36: Error: Unknown opcode: `stws|stw} %r21,-16(%sp)' {standard input}:39: Error: Invalid operands @ 1.7 log @more status update - many things work now - alpha kernels build - the rest of arm works - ia64 works as well as with gcc 12 - all the arm, m68k, mips, ppc, and sh3 kernels build - mips has some libstdc++ build issues, and some o32 atomic issues - sparc atf works @ text @d19 1 a19 1 aarch64eb y ? y y ? ? ? ? ? d38 1 a38 1 m68000 y ? y y n[2] ? ? ? ? d49 1 a49 1 sparc64 y y y y y y y ? ? d51 1 a51 1 x86_64 y y y y y ? y ? ? d80 3 d134 1 a134 1 earmv4: k k k y y k k k @ 1.6 log @update status for GCC 14 ports - most earm*, and powerpc64 now build - vax and ia64 have the sanitizer_asm.h problem. - some more port kernels build @ text @d20 1 a20 1 alpha y ? y y n[2] ? ? ? ? d22 1 a22 1 earmv4eb y ? y y ? ? ? ? ? d29 1 a29 1 earmv6hf y ? y y ? y ? ? ? d32 1 a32 1 earmv7eb y ? y y ? ? ? ? ? d37 1 a37 1 ia64 y ? y y n[2] ? ? ? ? d39 1 a39 1 m68k y ? y y n[2] ? ? ? ? d41 1 a41 1 mipsel y ? y y ? ? ? ? ? d46 3 a48 3 sh3eb y ? y y n[2] ? ? ? ? sh3el y ? y y n[2] ? ? ? ? sparc y y y y y y y ? ? d52 1 a52 1 riscv32 y b y y y y ? ? ? d72 8 a82 2 [X]: riscv32 kernels are not really implemented yet, not GCC 12 issue. [X]: GENERIC.SKI boots just as well as GCC 10. d131 3 a133 3 earmv4: ? ? ? ? ? ? ? y earm: y y ? ? earmhf: y ? d135 2 a136 2 earmv6hf: ? y earmv7: y ? d140 1 a140 1 m68k: ? ? ? ? ? ? ? ? ? ? ? d143 1 a143 1 mipseb: ? ? ? ? ? ? d146 1 a146 1 mipsel: ? ? ? ? ? ? d150 1 a150 1 mips64el: ? y ? d153 1 a153 1 powerpc: k k r r y r r r r r d156 1 a156 1 powerpc64: y ? ? d159 2 a160 2 sh3eb: ? ? sh3el: r ? r r @ 1.5 log @various updates to status - alpha, m68k, sh3, at least, have sanitizer_asm.h issues. sparc may be wrong here, since it seems to emit a single 'b func' with no delay slot. needs to be checked out for most CPUs fairly closely. - most of the arm ports build, have not run them yet - hppa has some codegen issue in libgcc (gas complains) - i386, mips64*, powerpc, and riscv64 builds now @ text @d23 4 a26 4 earmv5 y ? y y ? ? ? ? ? earmv5eb y ? y y ? ? ? ? ? earmv5hf y ? y y ? ? ? ? ? earmv5hfeb y ? y y ? ? ? ? ? d30 1 a30 1 earmv6hfeb y ? y y ? ? ? ? ? d37 1 a37 1 ia64 y ? y y ? r ? ? ? d45 1 a45 1 powerpc64 y ? y y y r ? ? ? d50 1 a50 1 vax y y[4] y y ? ? ? ? ? d126 2 a127 2 earm: ? ? ? ? earmhf: ? ? d133 1 a133 1 amiga atari cesfic hp300 luna68? mac68? mvme68? news68? next68? sun3 x68? d143 1 a143 1 mips64eb: ? d147 1 a147 1 powerpc: ? ? ? ? y ? ? ? ? ? d150 1 a150 1 powerpc64: ? ? ? d154 1 a154 1 sh3el: ? ? ? ? @ 1.4 log @update for more GCC 14 parts. armv4 builds, but needs the sanitizer treatment. sparc and sparc64 boot in qemu. @ text @d9 1 a9 1 sani: has sanitizer been ported to this CPU again for GCC 12? b (build) y (works) d20 2 a21 2 alpha y ? y y ? ? ? ? ? earmv4 y y y y y y ? ? ? d27 3 a29 3 earmv6 y ? y y ? ? ? ? ? earmv6eb y ? y y ? ? ? ? ? earmv6hf y ? y y ? ? ? ? ? d31 1 a31 1 earmv7 y ? y y ? ? ? ? ? d33 7 a39 7 earmv7hf y ? y y ? ? ? ? ? earmv7hfeb y ? y y ? ? ? ? ? hppa y ? y y ? ? ? ? ? i386 y ? y y ? ? ? ? ? ia64 y ? y y ? ? ? ? ? m68000 y ? y y ? ? ? ? ? m68k y ? y y ? ? ? ? ? d42 6 a47 6 mips64eb y ? y y ? ? ? ? ? mips64el y ? y y ? ? ? ? ? powerpc y ? y y ? ? ? ? ? powerpc64 y ? y y ? ? ? ? ? sh3eb y ? y y ? ? ? ? ? sh3el y ? y y ? ? ? ? ? d50 1 a50 1 vax y ? y y ? ? ? ? ? d52 2 a53 2 riscv32 y ? y y ? ? ? ? ? riscv64 y ? y y ? ? ? ? ? d58 15 d75 3 a77 3 [2]: riscv32 kernels are not really implemented yet, not GCC 12 issue. [3]: GENERIC.SKI boots just as well as GCC 10. [13]: on amd64, with very basic test, lsan, ubsan, and asan work, but tsan crashes at start up: d85 1 a85 1 [14]: "dd count=1", "env LC_CTYPE=en_US.UTF-8 locale", and "tests/crypto/libcrypto/h_evp_test evpmd_blake.txt" abort. d88 1 a88 1 [16]: sanitizers crash early: d95 1 a95 1 [17]: (NOT A REGRESSION) libtsan is newly introduced, which fails with: d97 1 a97 1 [23]: sanitizers fail to build: d99 1 a99 1 [24]: (NOT A REGRESSION) c++ exception handling is broken. mechanically applying d104 1 a104 1 [25]: (maybe not related to [24]) native gcc is broken: d109 1 a109 1 [26]: divide by zero with ubsan hits CheckLocked assertion failure, and stalls at LockSlow indefinitely: d128 4 a131 4 earmv6: ? ? earmv6hf: ? ? earmv7: ? ? earmv7hf: ? ? d144 1 a144 1 mips64el: ? ? ? d147 1 a147 1 powerpc: ? ? ? ? ? ? ? ? ? ? @ 1.3 log @note that sparc and sparc64 release build (untested yet), and there's an arm64 CTF issue @ text @d21 1 a21 1 earmv4 y ? y y ? ? ? ? ? d48 2 a49 2 sparc y ? y y y y ? ? ? sparc64 y ? y y y y ? ? ? d110 1 a110 1 earmv4: ? ? ? ? ? ? ? ? @ 1.2 log @update. mostly most ports mknative and libgcc fine. amd64 seems to work, some others almost build. @ text @d18 1 a18 1 aarch64 y ? y y ? ? ? ? ? d48 2 a49 2 sparc y ? y y ? ? ? ? ? sparc64 y ? y y ? ? ? ? ? d57 2 @ 1.1 log @various updates for GCC 14. - move README.gcc12 to gcc14 and clear it for fresh data - bump various shlib versions - add missing CXXFLAGS - add support for new/changed build things (options-urls.cc, genmatch and genemit generators) - fix up paths and some sources that aren't automatically handled - error->warning disables @ text @d18 36 a53 36 aarch64 y ? ? ? ? ? ? ? ? aarch64eb y ? ? ? ? ? ? ? ? alpha y ? ? ? ? ? ? ? ? earmv4 y ? ? ? ? ? ? ? ? earmv4eb y ? ? ? ? ? ? ? ? earmv5 y ? ? ? ? ? ? ? ? earmv5eb y ? ? ? ? ? ? ? ? earmv5hf y ? ? ? ? ? ? ? ? earmv5hfeb y ? ? ? ? ? ? ? ? earmv6 y ? ? ? ? ? ? ? ? earmv6eb y ? ? ? ? ? ? ? ? earmv6hf y ? ? ? ? ? ? ? ? earmv6hfeb y ? ? ? ? ? ? ? ? earmv7 y ? ? ? ? ? ? ? ? earmv7eb y ? ? ? ? ? ? ? ? earmv7hf y ? ? ? ? ? ? ? ? earmv7hfeb y ? ? ? ? ? ? ? ? hppa y ? ? ? ? ? ? ? ? i386 y ? ? ? ? ? ? ? ? ia64 y ? ? ? ? ? ? ? ? m68000 y ? ? ? ? ? ? ? ? m68k y ? ? ? ? ? ? ? ? mipseb y ? ? ? ? ? ? ? ? mipsel y ? ? ? ? ? ? ? ? mips64eb y ? ? ? ? ? ? ? ? mips64el y ? ? ? ? ? ? ? ? powerpc y ? ? ? ? ? ? ? ? powerpc64 y ? ? ? ? ? ? ? ? sh3eb y ? ? ? ? ? ? ? ? sh3el y ? ? ? ? ? ? ? ? sparc y ? ? ? ? ? ? ? ? sparc64 y ? ? ? ? ? ? ? ? vax y ? ? ? ? ? ? ? ? x86_64 y ? ? ? ? ? ? ? ? riscv32 y ? ? ? ? ? ? ? ? riscv64 y ? ? ? ? ? ? ? ? @