summaryrefslogtreecommitdiffstats
path: root/distrib
AgeCommit message (Collapse)AuthorFilesLines
2021-03-31synctb1-4/+4
2021-03-29Install apm(8) and and apmd(8) under MI MANPATHkn1-12/+2
powerpc64 was lacking manual pages; instead of shipping yet another identical MD manual, merge them under MI MANPATH as usual. Input OK deraadt
2021-03-26Resync the supported hardware list with armv7.html, suggested by,dtucker1-39/+28
corrections and ok jsg@. This replaces the obsolete list of boards and their supported devices with the current list of supported boards only, similar to arm64.
2021-03-24ensure the long-name version of ld is in the base set (rather than thederaadt29-15/+15
comp set), as it is supposed to be hard-linked. found by Yoshihiro Kawamata
2021-03-20SKIP_PROPOSAL has been ripped out in 2019kn1-3/+1
2021-03-19Remove libLLVM.so.2.0 on upgradekn1-2/+2
It is over a year old and corresponds to LLVM 8.0.0 after the "-msvr4-struct-return" ABI change; saves 47M (on amd64). OK deraadt
2021-03-15spellingjsg16-32/+32
2021-03-12syncderaadt1-0/+1
2021-03-12Emulate "[inet] autoconf" hostname.if(5) lines with "dhcp"kn1-1/+9
With dhcpleased(8) in base, netstart(8) and ifconfig(8) understand both "autoconf" and "inet autoconf" lines in hostname.if(5) files to signal the new daemon. The installer however currently has only dhclient(8), hence manual upgrades with "[inet] autoconf" instead of "dhcp" in hostname.if files would fail to establish IPv4 connectivity. Make install.sub's netstart clone treat autoconf lines like old fashioned dhcp lines such users^Wearly testers of the new approach don't get stuck in nyetwork land. Note that this is only relevant for manual upgrades; installation always creates working hostname.if files and automated upgrades with sysupgrade(8) do not care about network/hostname.if files. Idea from deraadt OK deraadt krw ajacoutot
2021-03-10Build install media with -fno-asynchronous-unwind-tables to furtherjsg1-1/+2
reduce size. Allows a clang 11 amd64 release to complete without overflowing the floppy image. ok kettenis@ deraadt@
2021-03-06syncderaadt1-1/+0
2021-03-01sync for perl 5.32.1afresh116-133/+52
2021-03-01syncderaadt1-0/+1
2021-03-01syncderaadt1-0/+1
2021-02-27syncderaadt1-0/+2
2021-02-26sync; OK deraadtflorian2-0/+6
2021-02-26syncderaadt1-0/+5
2021-02-24syncderaadt2-0/+2
2021-02-24syncderaadt1-1/+0
2021-02-24syncderaadt1-0/+1
2021-02-23syncderaadt1-14/+0
2021-02-23syncderaadt1-0/+1
2021-02-23syncderaadt1-0/+1
2021-02-21syncderaadt1-0/+1
2021-02-20syncsthen1-1/+1
2021-02-18Mention that U-Boot serial speed should be 115200.visa1-1/+7
Initial diff from Janne Johansson
2021-02-16make use of getline(3) in ftp(1)naddy1-4/+1
Replace fparseln(3) with getline(3). This removes the only use of libutil.a(fparseln.o) from the ramdisk. Replace a complicated fgetln(3) idiom with the much simpler getline(3). ok jca@
2021-02-16syncderaadt1-3/+0
2021-02-15distrib: arm64: avoid customization in mr.fs targetsemarie2-9/+4
arm64 ramdisk has customization in mr.fs target, in order to create usr/mdec/pine64 and usr/mdec/rpi directories (files will be copied inside them by runlist.sh). uses MKDIR directive to create these directories instead of. it put back mr.fs target identical to others archs. ok deraadt@
2021-02-15Things are too tight. Retaining even those two symbols is too much.deraadt1-2/+1
When dealing with install floppies, the Makefiles will need to carry some differences.
2021-02-14distrib: restore rdsetroot -x usage (extract the disk.fs image) on stripped bsd.rdsemarie9-19/+37
It passes options to keep rd_root_size and rd_root_image symbols while stripping. These symbols are the ones used by rdsetroot to insert or extract disk image into RAMDISK. ok danj@ deraadt@
2021-02-14distrib: use ${MACHINE} instead of hardcoded valuesemarie8-50/+50
ok deraadt@
2021-02-13distrib: remove .comment section systematicallysemarie4-8/+4
remove this section as part of crunchgen generated commands. it avoids calling "strip -R .comment" for some but not all architectures. ok deraadt@ danj@
2021-02-13distrib: remove .SUNW_ctf only on kernelssemarie7-18/+18
the .SUNW_ctf section is added by ctfstrip(1), which is only used for kernels. ok deraadt@ danj@
2021-02-13distrib: merge "cp a b; strip b" to "objcopy -S a b"semarie10-38/+24
ok deraadt@ danj@
2021-02-13distrib: remove duplicate strip(1) callssemarie6-16/+6
"strip -R section" command run "strip" (without option) as well. there is no need to call both. (binaries checked with sha1) original diff from danj@ ok deraadt@
2021-02-13distrib: merge objcopy -Sg and strip commandssemarie2-6/+4
in the command "objcopy -Sg", the -g option (STRIP_DEBUG) overrides the -S option (STRIP_ALL). so it is the same as "objcopy -g". "strip" command without option is doing STRIP_ALL. merge the both commands to "objcopy -S" ok deraadt@ danj@
2021-02-13distrib: remove STRIPOPTS variable from alpha and sparc64 Makefilesemarie2-9/+5
ok deraadt@ danj@
2021-02-12syncderaadt1-4/+0
2021-02-11syncderaadt7-21/+0
2021-02-09syncpatrick10-0/+30
2021-02-09syncderaadt1-0/+2
2021-02-09syncderaadt1-0/+1
2021-02-08syncderaadt1-0/+1
2021-02-07syncderaadt14-225/+225
2021-02-05syncderaadt1-0/+1
2021-02-04syncpatrick1-1/+0
2021-02-04syncanton1-0/+1
2021-02-01syncderaadt1-6/+0
2021-01-30Add dhclient.conf back to list of "installed network configurationkrw1-2/+2
files during upgrade". Mistakenly removed during dhclient.conf cleanup of r1.1050.
Copyright © 1996 – 2026 Jason A. Donenfeld. All Rights Reverse Engineered.