HACKS revision 1.61
11.61Smrg# $NetBSD: HACKS,v 1.61 2006/05/12 03:26:54 mrg Exp $ 21.1Slukem# 31.1Slukem# This file is intended to document workarounds for currently unsolved 41.1Slukem# (mostly) compiler bugs. 51.1Slukem# 61.1Slukem# Format: 71.1Slukem# hack title 81.1Slukem# cdate creation date 91.1Slukem# mdate mod date 101.1Slukem# who responsible developer 111.1Slukem# port ... 121.1Slukem# affected ports, space separated, if not "all" 131.1Slukem# file affected file : revision : line from : line to 141.1Slukem# affected files and revision and line numbers describing hack 151.1Slukem# multiple lines if necessary. 161.1Slukem# pr NNNN ... 171.1Slukem# problem reports this hack works around, if known. Space 181.1Slukem# separated. 191.1Slukem# regress src/regress/directory/where/test/found 201.1Slukem# regression test directories, if available. 211.1Slukem# descr 221.1Slukem# insert short informal description (multi-line). (Longer ones 231.1Slukem# should be in the PR database. More formal descriptions might 241.1Slukem# be in the regress tree. See above). 251.1Slukem# kcah 261.1Slukem# closing bracket. 271.1Slukem# 281.1Slukem# this is a comment. 291.1Slukem 301.1Slukemhack netstat ieee1394 address printing. 311.1Slukemmdate 14 Nov 2000 321.1Slukemwho matt 331.1Slukemfile lib/libc/net/getnameinfo.c : 1.32 : 497 : 503 341.1Slukemdescr 351.1Slukem Because the current implementation of IP over IEEE1394, the 361.1Slukem fw device address contains more than just the IEEE1394 EUI-64. 371.1Slukem So when printing out IEEE1394 addresses, ignore the extra stuff. 381.1Slukemkcah 391.1Slukem 401.1Slukemhack xterm vs. libterm 411.1Slukemmdate 01 Aug 2000 421.1Slukemwho jdc 431.1Slukemfile xsrc/xc/programs/xterm/main.c : 1.2 : 3609 : 3614 441.1Slukempr 10383 451.1Slukemdescr 461.1Slukem In order to extend the termcap string over 1023 bytes, a ZZ entry was 471.1Slukem introduced to point to a memory location containing the full entry. 481.1Slukem Without this hack, xterm will export a termcap containing the ZZ 491.1Slukem entry, which will then be ignored by libterm. As xterm modifies the 501.1Slukem exported termcap, this would cause those modifications to be ignored. 511.1Slukemkcah 521.1Slukem 531.1Slukemhack wi-at-big-endian-bus 541.1Slukemcdate 15 Mar 2002 551.1Slukemwho martin 561.1Slukemfile dev/ic/wireg.h 571.1Slukemdescr Add an option to access the underlying bus in big endian byte order 581.1Slukem to work around deficiencies in bus_space_{read,write}_* macros. 591.1Slukem Those don't allow the implementation of a proper pcmcia bus space 601.1Slukem tag. 611.1Slukemkcah 621.1Slukem 631.3Sjdchack specific knowledge of colours in curses code 641.3Sjdccdate Sun Apr 6 11:05:24 BST 2003 651.3Sjdcwho jdc 661.3Sjdcfile lib/libcurses/color.c : r1.24 671.3Sjdcdescr 681.3Sjdc Swap red/blue and yellow/cyan colours for COLOR_OTHER. 691.3Sjdc Fix is to enhance libtermcap to understand terminfo-style % sequences. 701.3Sjdc See also: 711.5Ssalo http://mail-index.NetBSD.org/tech-userlevel/2003/04/06/0000.html 721.4Sscwkcah 731.4Sscw 741.7Shehack Compensation for differing types of LINUX_USRSTACK and USRSTACK 751.7Shecdate 21 Aug 2003 761.7Shewho he 771.7Shefile sys/miscfs/procfs/procfs_linux.c : 1.14 781.7Shedescr 791.7She Not all ports have LINUX_USRSTACK and/or USRSTACK defined as 801.7She literals/constants, but refer to variables of a type which is 811.7She not "unsigned long", causing compilation of procfs_linux.c to 821.7She fail with "makes integer from pointer without a cast". This 831.7She is observed on e.g. the sun3 port. Ideally the "types" for 841.7She symbols should be consistent across all ports. 851.25Smrgkcah 861.25Smrg 871.25Smrghack GCC2 isn't C99 881.25Smrgcdate 5 December 2003 891.25Smrgwho mrg 901.25Smrgfile gnu/dist/binutils/ld/ldlex.c: r1.2 911.25Smrgdescr 921.25Smrg ldlex.c is generated from ldlex.l and we get macro redefinition 931.25Smrg errors from it if __STDC_VERSION__ is not C99 compliant. This 941.25Smrg hack forces NetBSD 1.6 and later systems to also include <stdint.h> 951.25Smrg to get the relevant definitions. This hack should be backed out 961.25Smrg when all platforms have switched away from GCC2. 971.11Slukemkcah 981.11Slukem 991.37Sjunyounghack gcc-strict-aliasing 1001.37Sjunyoungcdate 6 August 2004 1011.37Sjunyoungwho junyoung 1021.37Sjunyoungpr 26516 1031.37Sjunyoungfile src/usr.sbin/racoon/racoon/Makefile : 1.21 1041.37Sjunyoungdescr 1051.37Sjunyoung GCC issues a warning "dereferencing type-punned pointer will break 1061.37Sjunyoung strict-aliasing rules" when compiling 1071.37Sjunyoung crypto/dist/kame/racoon/crypto/openssl.c with -O[23s]. Note that this 1081.38Sjunyoung is due to use of type casting in a non-standard way in the code and 1091.38Sjunyoung not a compiler bug. Once a new release of KAME with a proper fix 1101.38Sjunyoung applied is imported this should be backed out. 1111.37Sjunyoungkcah 1121.37Sjunyoung 1131.58Smrghack gcc4 wrong uninitialised variable 1141.58Smrgmdate 10 May 2006 1151.58Smrgwho mrg 1161.58Smrgfile bin/ksh/eval.c : 1.6 1171.58Smrgfile bin/sh/histedit.c : 1.39 1181.58Smrgfile bin/sh/parser.c : 1.60 1191.58Smrgfile bin/systrace/systrace.c : 1.33 1201.58Smrgfile crypto/dist/heimdal/kdc/524.c : 1.10 1211.58Smrgfile crypto/dist/ssh/sftp.c : 1.20 1221.58Smrgfile crypto/dist/ssh/ssh-keysign.c : 1.11 1231.58Smrgfile dist/ipf/lib/hostname.c : 1.2 1241.58Smrgfile dist/ipf/tools/ipmon.c : 1.8 1251.58Smrgfile dist/ntp/ntpd/refclock_shm.c : 1.4 1261.58Smrgfile dist/ntp/sntp/timing.c : 1.3 1271.58Smrgfile dist/pppd/pppstats/pppstats.c : 1.3 1281.58Smrgfile dist/smbfs/lib/smb/rap.c : 1.6 1291.58Smrgfile dist/tcpdump/print-zephyr.c : 1.5 1301.58Smrgfile gnu/libexec/uucp/uucico/uucico.c : 1.6 1311.58Smrgfile lib/libc/citrus/citrus_csmapper.c : 1.6 1321.58Smrgfile lib/libc/citrus/citrus_pivot_factory.c : 1.5 1331.58Smrgfile lib/libc/inet/inet_cidr_ntop.c : 1.3 1341.58Smrgfile lib/libc/inet/inet_ntop.c : 1.3 1351.58Smrgfile lib/libc/stdio/vfwprintf.c : 1.8 1361.58Smrgfile libexec/ld.elf_so/arch/sparc64/mdreloc.c : 1.39 1371.58Smrgfile sbin/fsck_ext2fs/dir.c : 1.19 1381.58Smrgfile sbin/routed/rtquery/rtquery.c : 1.18 1391.58Smrgfile sys/crypto/cast128/cast128.c : 1.9 1401.58Smrgfile sys/ddb/db_command.c : 1.86 1411.60Smrgfile sys/dev/cardbus/cardbus_map.c : 1.xxx 1421.58Smrgfile sys/dev/fss.c : 1.25 1431.60Smrgfile sys/dev/ic/mb86950.c : 1.xxx 1441.58Smrgfile sys/dev/ic/midway.c : 1.71 1451.60Smrgfile sys/dev/usb/ohci.c : 1.xxx 1461.60Smrgfile sys/dev/usb/uhci.c : 1.xxx 1471.60Smrgfile sys/dev/pci/cmpci.c : 1.xxx 1481.58Smrgfile sys/dev/pci/machfb.c : 1.45 1491.58Smrgfile sys/dist/ipf/netinet/ip_nat.c : 1.10 1501.58Smrgfile sys/dist/ipf/netinet/ip_rpcb_pxy.c : 1.8 1511.58Smrgfile sys/dist/pf/net/pf.c : 1.22 1521.59Smrgfile sys/fs/udf/udf_vnops.c : 1.4 1531.58Smrgfile sys/kern/kern_sig.c : 1.219 1541.58Smrgfile sys/kern/tty.c : 1.181 1551.58Smrgfile sys/net/bpf.c : 1.116 1561.58Smrgfile sys/net/zlib.c : 1.26 1571.60Smrgfile sys/netccitt/if_x25subr.c : 1.xxx 1581.58Smrgfile sys/netinet/in.c : 1.107 1591.58Smrgfile sys/netsmb/smb_smb.c : 1.27 1601.58Smrgfile sys/netsmb/smb_trantcp.c : 1.22 1611.58Smrgfile sys/nfs/nfs_serv.c : 1.108 1621.58Smrgfile sys/nfs/nfs_socket.c : 1.129 1631.58Smrgfile sys/nfs/nfs_syscalls.c : 1.91 1641.58Smrgfile sys/ufs/lfs/lfs_vfsops.c : 1.207 [also (char *)] 1651.58Smrgfile usr.bin/ftp/ftp.c : 1.140 1661.58Smrgfile usr.bin/find/function.c : 1.54 1671.58Smrgfile usr.bin/mail/tty.c : 1.20 1681.58Smrgfile usr.bin/nl/nl.c : 1.7 1691.58Smrgfile usr.bin/systat/keyboard.c : 1.23 1701.58Smrgfile usr.bin/usbhidctl/usbhid.c : 1.29 1711.58Smrgfile usr.bin/vi/cl/cl_read.c : 1.5 1721.58Smrgfile usr.bin/vi/ex/ex_cscope.c : 1.12 1731.58Smrgfile usr.bin/vi/ex/ex_tag.c : 1.19 1741.58Smrgfile usr.bin/vi/vi/v_txt.c : 1.15 1751.58Smrgfile usr.sbin/altq/altqstat/qdisc_rio.c : 1.4 1761.58Smrgfile usr.sbin/cron/do_command.c : 1.19 1771.58Smrgfile usr.sbin/timed/timed/slave.c : 1.15 1781.58Smrgdescr 1791.58Smrg GCC 4.1 gets many uninitialised variable warnings wrong. We should 1801.58Smrg really audit all the old hacks like this when older compilers are 1811.58Smrg removed from the tree, as many are probably no longer required. 1821.58Smrgkcah 1831.58Smrg 1841.59Smrghack gcc4 pointer sign and strict aliasing problems 1851.59Smrgmdate 10 May 2006 1861.59Smrgwho mrg 1871.59Smrgfile bin/ed/Makefile : 1.33 1881.60Smrgfile bin/systrace/Makefile : 1.12 1891.60Smrgfile games/bcd/Makefile : 1.5 1901.60Smrgfile games/dab/Makefile : 1.5 1911.60Smrgfile games/larn/Makefile : 1.17 1921.60Smrgfile games/pom/Makefile : 1.5 1931.60Smrgfile lib/libasn1/Makefile : 1.26 1941.60Smrgfile lib/libcrypt/Makefile : 1.17 1951.60Smrgfile lib/libgssapi/Makefile : 1.16 1961.60Smrgfile lib/libhdb/Makefile : 1.20 1971.60Smrgfile lib/libkadm5clnt/Makefile : 1.21 1981.60Smrgfile lib/libkadm5srv/Makefile : 1.25 1991.60Smrgfile lib/libkrb5/Makefile : 1.35 2001.60Smrgfile lib/libssh/Makefile : 1.6 2011.60Smrgfile lib/libtelnet/Makefile : 1.26 2021.60Smrgfile libexec/getty/Makefile : 1.14 2031.60Smrgfile libexec/kadmind/Makefile : 1.19 2041.60Smrgfile libexec/kpasswdd/Makefile : 1.14 2051.60Smrgfile sbin/atactl/Makefile : 1.3 2061.60Smrgfile sbin/cgdconfig/Makefile : 1.7 2071.60Smrgfile sbin/clri/Makefile : 1.13 2081.60Smrgfile sbin/dkctl/Makefile : 1.4 2091.60Smrgfile sbin/dump/Makefile : 1.33 2101.60Smrgfile sbin/fdisk/Makefile : 1.35 2111.60Smrgfile sbin/fsck_ext2fs/Makefile : 1.11 2121.60Smrgfile sbin/fsck_ffs/Makefile : 1.29 2131.60Smrgfile sbin/fsdb/Makefile : 1.18 2141.60Smrgfile sbin/mount_smbfs/Makefile : 1.4 2151.60Smrgfile sbin/newfs/Makefile : 1.30 2161.60Smrgfile sbin/newfs_sysvbfs/Makefile : 1.2 2171.60Smrgfile sbin/restore/Makefile : 1.23 2181.60Smrgfile sbin/veriexecctl/Makefile : 1.11 2191.59Smrgfile usr.bin/awk/Makefile : 1.9 2201.60Smrgfile usr.bin/ctags/Makefile : 1.8 2211.61Smrgfile usr.bin/gzip/Makefile : 1.10 2221.60Smrgfile usr.bin/ssh/sftp/Makefile : 1.10 2231.60Smrgfile usr.bin/ssh/ssh/Makefile : 1.25 2241.59Smrgfile usr.bin/vi/build/Makefile : 1.26 2251.60Smrgfile usr.bin/telnet/Makefile : 1.40 2261.60Smrgfile usr.bin/tn3270/tn3270/Makefile : 1.36 2271.60Smrgfile usr.bin/tr/Makefile : 1.4 2281.59Smrgfile usr.sbin/amd/amd/Makefile : 1.27 2291.59Smrgfile usr.sbin/amd/amq/Makefile : 1.14 2301.59Smrgfile usr.sbin/amd/libamu/Makefile : 1.20 2311.59Smrgfile usr.sbin/amd/pawd/Makefile : 1.5 2321.59Smrgfile usr.sbin/bind/Makefile.inc : 1.22 2331.60Smrgfile usr.sbin/bind/libdns/Makefile : 1.3 2341.60Smrgfile usr.sbin/bind/named/Makefile : 1.17 2351.60Smrgfile usr.sbin/bootp/bootptest/Makefile : 1.2 2361.60Smrgfile usr.sbin/chrtbl/Makefile : 1.6 2371.60Smrgfile usr.sbin/cron/Makefile : 1.12 2381.59Smrgfile usr.sbin/dhcp/Makefile.inc : 1.20 2391.59Smrgfile usr.sbin/dumplfs/Makefile : 1.11 2401.60Smrgfile usr.sbin/hprop/Makefile : 1.13 2411.60Smrgfile usr.sbin/installboot/Makefile : 1.35 2421.59Smrgfile usr.sbin/ipf/ipftest/Makefile : 1.32 2431.59Smrgfile usr.sbin/isdn/isdnd/Makefile : 1.6 2441.59Smrgfile usr.sbin/isdn/isdnmonitor/Makefile : 1.3 2451.59Smrgfile usr.sbin/isdn/isdntel/Makefile : 1.2 2461.59Smrgfile usr.sbin/isdn/isdntrace/Makefile : 126 2471.61Smrgfile usr.sbin/mopd/common/Makefile : 1.10 2481.61Smrgfile usr.sbin/mopd/mopd/Makefile : 1.9 2491.61Smrgfile usr.sbin/mopd/mopprobe/Makefile : 1.7 2501.60Smrgfile usr.sbin/makefs/Makefile : 1.17 2511.60Smrgfile usr.sbin/mscdlabel/Makefile : 1.5 2521.60Smrgfile usr.sbin/pkg_install/add/Makefile : 1.7 2531.60Smrgfile usr.sbin/pkg_install/create/Makefile : 1.5 2541.60Smrgfile usr.sbin/pkg_install/lib/Makefile : 1.28 2551.59Smrgfile usr.sbin/ntp/ntpd/Makefile : 1.10/1.11 2561.59Smrgfile usr.sbin/ntp/ntptime/Makefile : 1.4/1.5 2571.59Smrgfile usr.sbin/pppd/Makefile.inc : 1.3 2581.59Smrgfile usr.sbin/pppd/pppd/Makefile : 1.38 2591.60Smrgfile usr.sbin/rarpd/Makefile : 1.10 2601.59Smrgfile usr.sbin/rbootd/Makefile : 1.10 2611.60Smrgfile usr.sbin/rpc.pcnfsd/Makefile : 1.17 2621.59Smrgfile usr.sbin/rtadvd/Makefile : 1.9 2631.60Smrgfile usr.sbin/tcpdump/Makefile : 1.42 2641.59Smrgfile usr.sbin/wiconfig/Makefile : 1.3 2651.60Smrgfile usr.sbin/ypserv/ypserv/Makefile : 1.16 2661.59Smrgdescr 2671.59Smrg GCC 4.1 warns on pointer sign comparision/assignments and lots of 2681.59Smrg code does not conform. For now we use -Wno-pointer-sign and 2691.59Smrg -fno-strict-aliasing. 2701.59Smrgkcah 2711.59Smrg 2721.13Ssimonb 2731.21Smycroftport vax 2741.21Smycroft 2751.52Sjmc hack gcc 2.95/vax doesn't like abort being used as function pointer 2761.52Sjmc cdate Tue Dec 13 05:54:50 GMT 2005 2771.52Sjmc who jmc 2781.52Sjmc file lib/libc/rpc/svc_vc.c 1.15 2791.52Sjmc descr 2801.52Sjmc Provide a local definition for abort which doesn't include 2811.52Sjmc the noreturn attribute which trips up gcc 2.95 on vax. 2821.52Sjmc 2831.45Sjmc hack gcc 2.95/vax doesn't have stdbool.h and gettext needs it 2841.45Sjmc cdate Mon May 9 12:16:22 CDT 2005 2851.45Sjmc who jmc 2861.45Sjmc file gnu/usr.bin/gettext/Makefile.inc.prog 1.5 2871.45Sjmc gnu/usr.bin/gettext/libgrep/Makefile 1.3 2881.45Sjmc gnu/usr.bin/gettext/libnlspr/Makefile 1.5 2891.45Sjmc gnu/usr.bin/gettext/libnlsut/Makefile 1.3 2901.45Sjmc descr 2911.45Sjmc The latest gettext assumes a C99 environment or at least 2921.45Sjmc configure checks to work around this. As we don't use configure 2931.45Sjmc provide a stdbool.h for vax builds locally in directories from 2941.45Sjmc the template provided by the gettext code. 2951.45Sjmc 2961.44Sjmc hack gcc 2.95/vax can't handle __used__ in mcount.c 2971.44Sjmc cdate Fri May 6 14:30:53 CDT 2005 2981.44Sjmc who jmc 2991.44Sjmc file src/lib/libc/gmon/mcount.c 1.17 3001.44Sjmc descr 3011.44Sjmc Recent changes to mark this as __used__ to prevent optimization 3021.44Sjmc away on sparc64 don't work on 2.95 vax. Ifdef'd back in the 3031.44Sjmc old __unused__ if on vax. 3041.44Sjmc 3051.21Smycroft hack gcc 2.95/vax cannot handle __builtin_ffs() 3061.21Smycroft cdate Fri Aug 23 21:31:15 CEST 2002 3071.21Smycroft who ragge 3081.21Smycroft file sys/lib/libkern/libkern.h : 1.42 3091.21Smycroft descr 3101.21Smycroft __builtin_ffs() is ifdef'd away if __vax__. Matt Thomas has 3111.21Smycroft added this feature to GCC 3 so it can be removed when switching. 3121.44Sjmc 3131.21Smycroft 3141.21Smycroft hack gcc-vax-libbz2 3151.21Smycroft mdate 27 Jun 2002 3161.21Smycroft who thorpej 3171.21Smycroft file lib/libbz2/Makefile 3181.21Smycroft descr 3191.21Smycroft libbz2 is mis-compiled with optimization with GCC 2.95.3 3201.24Sragge on VAX. -O0 works around this problem. 3211.24Sragge kcah 3221.24Sragge 3231.24Sragge hack gawk 3241.24Sragge mdate 8 Nov 2003 3251.24Sragge who ragge 3261.24Sragge file gnu/usr.bin/gawk/Makefile 3271.24Sragge descr 3281.24Sragge gawk is mis-compiled with optimization with GCC 2.95.3 3291.21Smycroft on VAX. -O0 works around this problem. 3301.21Smycroft kcah 3311.21Smycroft 3321.21Smycroft hack vax Toolchain bug 3331.21Smycroft cdate 28 Mar 2003 3341.21Smycroft who he 3351.21Smycroft file sys/arch/vax/vax/intvec.S : 1.5 3361.21Smycroft descr 3371.21Smycroft Workaround for PR toolchain/20924. The assembler apparently 3381.21Smycroft tries to range-check byte offsets when it doesn't have 3391.21Smycroft sufficient information to make that decision. 3401.21Smycroft Workaround uses `brw' instead of `brb' instruction. 3411.23She kcah 3421.23She 3431.23She hack vax gcc 2.95.3 -Wuninitialized workarounds 3441.23She cdate 6 Nov 2003 3451.23She who he 3461.23She file sys/dev/bi/if_ni.c : 1.22 3471.23She file sys/arch/vax/vax/ctu.c : 1.20 3481.23She descr 3491.23She Workaround for gcc 2.95.3 failing to detect that certain 3501.23She variables will be initialized; so that this code compiles 3511.23She with -Wuninitialized. Should be reviewed and possibly 3521.23She reverted when gcc 3.3.2 is ready for vax. 3531.21Smycroft kcah 3541.21Smycroft 3551.34Scl hack vax gcc 2.95.3 structure initialization 3561.34Scl cdat 13 May 2004 3571.34Scl who cl 3581.34Scl file sys/miscfs/kernfs/kernfs_vnops.c : 1.104 3591.34Scl descr 3601.34Scl Workaround for gcc 2.95.3 failing to initialize structures 3611.34Scl and/or unions inside structures using nested designators. 3621.34Scl Should be reverted when gcc >=3.3.3 is ready for vax. 3631.34Scl kcah 3641.21Smycroft 3651.35She hack vax gcc 2.95.3 needs -I. to build nslexer.c 3661.35She cdat 6 Jun 2004 3671.36She who mhitch 3681.36She file lib/libc/arch/vax/Makefile.inc : 1.6 3691.35She descr 3701.35She It appears necessary to add -I. to CPPFLAGS when building 3711.35She nslesxer.c. This may be caused by gcc 2.95.3 being used. 3721.35She Should be reverted when a newer gcc is ready for vax. 3731.35She kcah 3741.35She 3751.43Stron hack vax gcc 2.95.3 doesn't understand "-std=c99" 3761.43Stron cdat 5 April 2005 3771.43Stron who tron 3781.43Stron file src/share/mk/bsd.sys.mk 3791.43Stron descr 3801.43Stron We cannot use "-std=c99" for compiler warning level 4 and above 3811.43Stron because gcc 2.95.3 doesn't support that option. 3821.43Stron kcah 3831.43Stron 3841.54She hack declare boolean_t in two IPF user-mode programs 3851.54She cdate Tue Mar 7 19:19:20 CET 2006 3861.54She who he 3871.54She file dist/ipf/ipsend/iptests.c : 1.8 3881.54She dist/ipf/ipsend/sock.c : 1.7 3891.54She descr 3901.54She The IPF user-mode programs ipsend and iptest first 3911.54She include <sys/types.h> without _KERNEL defined, and 3921.54She later include <sys/file.h> with _KERNEL defined. 3931.54She This causes a build failure when building for vax, 3941.54She since <sys/device.h> ends up being included without 3951.54She bollean_t being defined by <sys/types.h>. 3961.54She Build failure and further details documented in 3971.54She PR#32907. 3981.54She kcah 3991.54She 4001.21Smycroftport sparc64 4011.21Smycroft 4021.21Smycroft hack binutil-2.11-sparc64-pltrela 4031.21Smycroft mdate 14 Aug 2001 4041.21Smycroft who eeh 4051.21Smycroft file libexec/ld.elf_so/arch/sparc64/mdreloc.c 4061.21Smycroft descr 4071.39Sskrll The first four PLT entries are reserved. There is some 4081.21Smycroft disagreement whether they should have associated relocation 4091.21Smycroft entries. Both the SPARC 32-bit and 64-bit ELF specifications 4101.21Smycroft say that they should have relocation entries, but the 32-bit 4111.21Smycroft SPARC binutils do not generate them, and now the 64-bit SPARC 4121.21Smycroft binutils have stopped generating them too. 4131.21Smycroft 4141.21Smycroft To provide binary compatibility, we will check the first entry, 4151.21Smycroft if it is reserved it should not be of the type JMP_SLOT. If it 4161.21Smycroft is JMP_SLOT, then the 4 reserved entries were not generated and 4171.21Smycroft our index is 4 entries too far, so we frob the rela pointer. 4181.21Smycroft kcah 4191.21Smycroft 4201.21Smycroft 4211.21Smycroftport arm 4221.21Smycroft 4231.21Smycroft hack gcc-unsigned-compare 4241.21Smycroft cdate 09 Mar 2002 4251.21Smycroft mdate 18 Mar 2002 4261.21Smycroft who bjh21 4271.21Smycroft file dist/bind/lib/nameser/ns_parse.c : 1.3 4281.21Smycroft file dist/dhcp/minires/ns_parse.c : 1.3 4291.21Smycroft file dist/dhcp/omapip/result.c : 1.2 4301.21Smycroft file dist/dhcp/server/failover.c : 1.3 4311.21Smycroft file gnu/dist/gawk/eval.c : 1.4 4321.21Smycroft file gnu/dist/toolchain/bfd/bfd.c : 1.2 4331.21Smycroft file gnu/dist/toolchain/bfd/format.c : 1.2 4341.21Smycroft file gnu/dist/toolchain/gdb/target.c : 1.2 4351.21Smycroft file sys/kern/vfs_subr.c : 1.172 4361.21Smycroft descr When checking that a potentially-unsigned enum is >= 0, assign 4371.21Smycroft it to an int first. This is necessary to avoid "comparison is 4381.21Smycroft always true" warnings with -fshort-enums. Casting to an int 4391.21Smycroft really should be enough, but turns out not to be. 4401.21Smycroft kcah 4411.21Smycroft 4421.21Smycroft hack gcc 3.3.2/arm sys/kern/uipc_socket.c 4431.21Smycroft cdate Tue Oct 28 18:02:16 GMT 2003 4441.21Smycroft who skrll 4451.41Sbjh21 pr 23044 4461.41Sbjh21 file sys/arch/arm/conf/Makefile.arm : 1.19 4471.21Smycroft descr 4481.21Smycroft Performing a build.sh tools on a CATS with NFS mounted 4491.21Smycroft obj and src directories results in "panic: receive 1" 4501.21Smycroft The hack is to compile sys/kern/uipc_socket.c with 4511.21Smycroft -fno-strict-aliasing 4521.21Smycroft kcah 4531.21Smycroft 4541.21Smycroft 4551.13Ssimonbport pc532 4561.16Scl 4571.21Smycroft hack egcs-pc532-ip6_mroute 4581.21Smycroft cdate 09 Jul 2002 4591.21Smycroft who simonb 4601.21Smycroft file sys/arch/pc532/conf/Makefile.pc532 : 1.70 4611.21Smycroft file sys/arch/pc532/conf/files.pc532 : 1.47 4621.21Smycroft descr 4631.21Smycroft egcs 1.1.2 gets an "internal error--insn does not satisfy its 4641.21Smycroft constraints" error compiling ip6_mroute.c with -O2 or greater. 4651.21Smycroft -O1 works around this problem. 4661.21Smycroft kcah 4671.21Smycroft 4681.21Smycroft hack gcc-pc532-duffs_device 4691.21Smycroft cdate Mon Oct 27 07:23:05 UTC 2003 4701.21Smycroft who simonb 4711.21Smycroft file sys/arch/pc532/conf/Makefile.pc532 : 1.71 4721.21Smycroft file sys/arch/pc532/conf/files.pc532 : 1.52 4731.21Smycroft pr GCC PR optimization/5230 4741.21Smycroft descr 4751.21Smycroft gcc incorrectly gives an "unreachable code at beginning of 4761.21Smycroft switch statement" for a Duff's device construct in 4771.21Smycroft arch/pc532/dev/ncr.c. There is no way to disable just this 4781.21Smycroft warning, so -Wno-error is turned on for this file. 4791.21Smycroft kcah 4801.21Smycroft 4811.26Ssimonb hack gcc332-cppmacro-ice 4821.26Ssimonb cdate Sun Dec 7 12:48:33 UTC 2003 4831.26Ssimonb who simonb 4841.26Ssimonb file gnu/usr.bin/gcc3/libcpp/Makefile : 1.2 4851.26Ssimonb descr 4861.26Ssimonb gcc 3.3.2 gets an ICE when compiling cppmacro.c with -O2. 4871.26Ssimonb The hack is to use -O1 for this file. Fixed in gcc 4881.26Ssimonb -current mainline, unable to work out where the fix is 4891.26Ssimonb right now. 4901.26Ssimonb kcah 4911.26Ssimonb 4921.21Smycroft 4931.21Smycroftport sh3 4941.21Smycroft 4951.21Smycroft hack gcc-sh3-sed 4961.21Smycroft mdate 23 Apr 2002 4971.21Smycroft who thorpej 4981.21Smycroft file usr.bin/sed/Makefile : 1.9 : 9 : 13 4991.21Smycroft descr 5001.21Smycroft The in-tree GCC 2.95.3-based compiler ICEs when building 5011.26Ssimonb with optimization for SuperH. Hack is to build with -O0. 5021.21Smycroft kcah 5031.21Smycroft 5041.21Smycroft 5051.21Smycroftport sh5 5061.21Smycroft 5071.21Smycroft hack SuperH SH5 Toolchain Bugs 5081.21Smycroft cdate 11 Jul 2002 5091.21Smycroft who scw 5101.21Smycroft file usr.sbin/ndbootd/ndbootd.c : 1.6 5111.21Smycroft file usr.sbin/traceroute/traceroute.c : 1.48 5121.21Smycroft descr 5131.21Smycroft The SuperH SH5 toolchain (2.97-sh5-010522) gets an internal 5141.21Smycroft compiler error when assigning a bit-wise inverted value 5151.21Smycroft under some circumstances. 5161.21Smycroft Work around it by splitting the statement into two. 5171.21Smycroft kcah 5181.21Smycroft 5191.21Smycroft hack SuperH SH5 Toolchain Bugs 5201.21Smycroft cdate 17 May 2003 5211.21Smycroft who scw 5221.21Smycroft file lib/libpthread/pthread_lock.c : 1.7 5231.21Smycroft descr 5241.21Smycroft The SuperH SH5 toolchain generates incorrect PIC code when 5251.21Smycroft faced with a symbol which is declared extern, but has local 5261.21Smycroft scope due to being defined within an asm() statement (without 5271.21Smycroft being declared .globl in that statement). Work around it by 5281.21Smycroft adding the .globl. 5291.27Sscw kcah 5301.27Sscw 5311.27Sscw hack gcc 3.3.2 on sh5 - ICE compiling reload1.c at -O2 5321.27Sscw cdate Sun Dec 7 19:44:05 GMT 2003 5331.27Sscw who scw 5341.27Sscw file gnu/usr.bin/gcc3/backend/Makefile : 1.13 5351.27Sscw descr 5361.27Sscw Gcc3 for sh5 gets an ICE compiling reload1.c at -O2. 5371.27Sscw Drop to -O1 for now when building for sh5. 5381.27Sscw kcah 5391.28Sscw 5401.28Sscw hack gcc 3.3.2 on sh5 - ICE compiling elf32.c/elf64.c at -O2 5411.28Sscw cdate Sun Dec 7 21:10:46 GMT 2003 5421.28Sscw who scw 5431.28Sscw file gnu/lib/libbfd/Makefile : 1.39 5441.28Sscw descr 5451.28Sscw Gcc3 for sh5 gets an ICE compiling libbfd's elf32.c 5461.28Sscw and elf64.c at -O2. 5471.28Sscw Drop to -O1 for now when building for sh5. 5481.28Sscw kcah 5491.28Sscw 5501.27Sscw 5511.27Sscw hack gcc 3.3.2 on sh5 - fatal warning compiling function.c 5521.27Sscw cdate Sun Dec 7 19:44:05 GMT 2003 5531.27Sscw who scw 5541.27Sscw file gnu/usr.bin/gcc3/backend/Makefile : 1.13 5551.27Sscw descr 5561.27Sscw Gcc3 for sh5 issues a warning: 5571.27Sscw .../gnu/dist/gcc/gcc/function.c:4742: warning: `and' of 5581.27Sscw mutually exclusive equal-tests is always 0. 5591.27Sscw This causes the build to fail because of -Werror. 5601.27Sscw Work-around by defining NOGCCERROR when building for sh5. 5611.27Sscw kcah 5621.27Sscw 5631.40She hack gcc 3.3.3 on sh5 - ICE compiling rbtdb.c and rbtdb64.c 5641.40She cdate Thu Nov 11 00:35:33 UTC 2004 5651.40She who he 5661.40She file usr.sbin/bind/libdns/Makefile/1.2 5671.27Sscw descr 5681.40She Gcc3 for sh5 gets an ICE compiling these files at -O2. 5691.27Sscw Drop to -O1 for now when building for sh5. 5701.21Smycroft kcah 5711.17Smrg 5721.48She hack gcc 3.3.3 nb3 on sh5 - ICE compiling bsd-comp.c 5731.48She cdate Thu May 25 23:05:20 UTC 2005 5741.48She who he 5751.48She file sys/lkm/net/bsdcomp/Makefile : 1.3 5761.48She descr 5771.48She Gcc3 for sh5 gets an ICE compiling this at -O2, 5781.48She "could not split insn". Drop to -O1 for sh5. 5791.48She kcah 5801.17Smrg 5811.53Sscw hack gcc 3.3.3 nb3 on sh5 - ICE compiling strtod.c 5821.53Sscw cdate Tue Jan 31 10:51:52 GMT 2006 5831.53Sscw who scw 5841.53Sscw file lib/libc/gdtoa/Makefile.inc : 1.2 5851.53Sscw descr 5861.53Sscw Gcc3 for sh5 gets an ICE compiling this at -O2, 5871.53Sscw "could not split insn". Drop to -O1 for sh5. 5881.53Sscw kcah 5891.53Sscw 5901.17Smrgport sun2 5911.20Sskrll 5921.21Smycroft hack gcc 3.3.2/mdsetimage 5931.21Smycroft cdate Tue Oct 28 18:43:05 EST 2003 5941.21Smycroft who fredette 5951.21Smycroft file gnu/usr.sbin/mdsetimage/Makefile : 1.15 5961.21Smycroft descr 5971.21Smycroft mdsetimage.c causes cc1 to barf for as-yet unknown reasons. 5981.26Ssimonb this hack uses -O0 to avoid the problem. 5991.21Smycroft kcah 6001.21Smycroft 6011.21Smycroft hack gcc 3.3.2/installboot/ffs.c 6021.21Smycroft cdate Tue Oct 28 18:43:05 EST 2003 6031.21Smycroft who fredette 6041.21Smycroft file usr.sbin/installboot/Makefile : 1.26 6051.21Smycroft descr 6061.21Smycroft ffs.c causes cc1 to barf for as-yet unknown reasons. this 6071.26Ssimonb hack uses -O0 to avoid the problem. 6081.22Smrg kcah 6091.22Smrg 6101.22Smrg 6111.56Stsutsuiport m68000 6121.22Smrg 6131.22Smrg hack gcc 3.3.2/gcc/ra-colorize.c 6141.22Smrg cdate Wed Nov 5 11:03:13 EST 2003 6151.22Smrg who mrg 6161.46Schs file gnu/usr.bin/gcc3/backend/Makefile : 1.11 6171.22Smrg descr 6181.22Smrg ra-colorize.c causes a whole spew of unreferenced .L* local 6191.22Smrg symbols. using -O0 (or even -O2 reportedly) avoids the problem. 6201.56Stsutsui Note this has been fixed in 3.3.3 at least on m68k, 6211.56Stsutsui but it hasn't been confirmed on sun2/m68000. 6221.56Stsutsui (it's too slow even on emulator and unlikely to build 6231.56Stsutsui toolchains on sun2 with its maximum 4MB RAM) 6241.21Smycroft kcah 6251.31Smartin 6261.31Smartinport sparc64 6271.31Smartin 6281.31Smartin hack disable optimzations for uvm_bio.c on 32 bit kernels 6291.31Smartin cdate Sun Mar 21 14:14:04 MET 2004 6301.31Smartin who martin 6311.32Smartin file sys/arch/sparc64/conf/Makefile.sparc64 : 1.53-1.54 6321.32Smartin file sys/arch/sparc64/conf/files.sparc64 : 1.93 6331.31Smartin descr 6341.31Smartin The sparc compiler miscompiles uvm_bio.c when using 6351.31Smartin any optimization. This results in ubc_release panics. 6361.31Smartin kcah 6371.42Schristos 6381.42Schristosport macppc 6391.42Schristos 6401.42Schristos hack remove inline from rotate_{left,right} 6411.42Schristos cdate Tue Feb 22 13:43:25 EST 2005 6421.42Schristos who christos 6431.42Schristos file src/dist/bind/lib/dns/rbt.c 6441.42Schristos descr 6451.42Schristos The macppc gcc-3.3 compiler miscompiles rbt.c when these 6461.42Schristos functions are inlined. The error is: 6471.42Schristos .* named[368]: rbt.c:1584: INSIST(child != (void *)0) >failed 6481.42Schristos .* named[368]: exiting (due to assertion failure) 6491.42Schristos This is reportedly fixed in gcc-3.4 6501.42Schristos kcah 6511.46Schs 6521.46Schsport hppa 6531.46Schs 6541.46Schs hack hppa-gcc-3.3.3/grep/dfa.c 6551.46Schs cdate Sun Aug 8 08:34:32 PDT 2004 6561.46Schs who chs 6571.46Schs file gnu/usr.bin/grep/grep/Makefile : 1.4 6581.46Schs descr 6591.46Schs gcc 3.3.3 generates bad code for dfa.c with -O2, 6601.46Schs reduce optimization to -O1 to avoid the bug. 6611.46Schs kcah 6621.47Schs 6631.55She hack hppa-gcc-3.3.3 / make/buf.c 6641.55She cdate Mon May 8 13:02:36 CEST 2006 6651.55She who he 6661.55She file usr.bin/make/Makefile : 1.34 6671.55She descr 6681.55She gcc 3.3.3 nb3 gets an "unrecognized insn" internal 6691.55She compiler error when building this file with -O2, 6701.55She reduce optimization to -O1 to avoid the bug. 6711.55She kcah 6721.55She 6731.47Schs hack hppa boot-from-disk memory corruption 6741.47Schs cdate Wed May 18 06:56:07 PDT 2005 6751.47Schs who chs 6761.47Schs file src/sys/dev/scsipi/scsipiconf.h : 1.98 6771.47Schs descr 6781.47Schs There is some bug with DMA on hppa that corrupts scsipi_xfer 6791.47Schs structures, but it only shows up when booting from disk. 6801.47Schs For now, we'll add some padding to scsipi_xfer so that 6811.47Schs the corrupted memory is not otherwise used. 6821.47Schs kcah 6831.51Ssimonb 6841.51Ssimonbport mips 6851.51Ssimonb 6861.51Ssimonb hack mips-shared-linker-load-address 6871.51Ssimonb cdate Fri Oct 7 08:33:10 UTC 2005 6881.51Ssimonb who simonb 6891.51Ssimonb file src/sys/kern/exec_elf32.c : 1.107 6901.51Ssimonb descr 6911.51Ssimonb With COMPAT_16 or previous enabled (which enables 6921.51Ssimonb ELF_INTERP_NON_RELOCATABLE) a recent ld.elf_so will 6931.51Ssimonb load and run at address 0. The check to fix this in 6941.51Ssimonb rev 1.107 only checks the first psection of the ELF 6951.51Ssimonb executable, which may not be loadable. A more correct 6961.51Ssimonb fix is to check the first loadable psection instead of 6971.51Ssimonb just the first psection. 6981.51Ssimonb kcah 6991.54She 700