OpenBSD src changes summary for 2016-01-10 to 2016-01-17 inclusive ==================================================================
bin/ksh bin/ps bin/test distrib/miniroot distrib/sets distrib/special games/arithmetic games/boggle games/bs games/canfield games/cribbage games/fish games/fortune games/hack games/hangman games/hunt games/phantasia games/quiz games/random games/sail games/tetris games/wump gnu gnu/usr.bin/perl regress/bin regress/usr.bin sbin/ifconfig sbin/isakmpd sbin/pdisk sbin/pfctl sbin/pflogd share/man share/mk sys/arch/alpha/pci sys/arch/amd64/amd64 sys/arch/amd64/conf sys/arch/amd64/include sys/arch/amd64/stand/libsa sys/arch/armish/conf sys/arch/armv7/exynos sys/arch/armv7/imx sys/arch/armv7/omap sys/arch/i386/conf sys/arch/i386/stand/libsa sys/arch/loongson/conf sys/arch/mips64/include sys/arch/mips64/mips64 sys/arch/octeon/conf sys/arch/octeon/dev sys/arch/octeon/octeon sys/arch/sh/include sys/arch/sh/sh sys/ddb sys/dev sys/dev/acpi sys/dev/i2c sys/dev/ic sys/dev/pci sys/dev/pv sys/dev/sdmmc sys/dev/usb sys/dev/wscons sys/kern sys/msdosfs sys/net sys/net80211 sys/netinet sys/netinet6 sys/tmpfs sys/ufs/ext2fs sys/ufs/ffs usr.bin/aucat usr.bin/calendar usr.bin/jot usr.bin/kdump usr.bin/less usr.bin/mandoc usr.bin/newsyslog usr.bin/rpcgen usr.bin/ssh usr.bin/systat usr.bin/tmux usr.bin/which usr.sbin/cron usr.sbin/dhcrelay usr.sbin/eigrpctl usr.sbin/eigrpd usr.sbin/hostapd usr.sbin/installboot usr.sbin/ldapd usr.sbin/lpr usr.sbin/ndp usr.sbin/ntpd usr.sbin/relayd usr.sbin/smtpd usr.sbin/sysmerge usr.sbin/tcpdump usr.sbin/vmctl usr.sbin/vmd == bin =============================================================== 01/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/bin ksh ~ c_ksh.c > Initialize a struct kill_info and an associated int in the declarations > instead of scattering their initialization through the body. This makes > the body simpler and avoids an uninitialized use false positive. > ok nicm@ (mmcc@) ps - fmt.c ~ Makefile ~ extern.h ~ print.c ~ ps.c + utf8.c > UTF-8 support: > In a UTF-8 locale, columnate correctly and replace valid, but non- > printable characters with the Unicode replacement character U+FFFD. > No change in the C/POSIX locale, and no change for invalid bytes. > Grand total, the code becomes shorter by almost 30 lines. > Feedback from czarkoff@, OK millert@. (schwarze@) + fmt.c > Partial revert: put back the file fmt.c. > It's no longer used by ps(1), but w(1) reaches around to it. > Cluestick applied by deraadt@. (schwarze@) test ~ test.c > fix exit status on pledge fail, eval here should be 2. > OK tb@ (gsoares@) == distrib =========================================================== 02/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/distrib miniroot ~ install.sub > simplify pattern > feedback halex@ > OK krw@ (rpe@) ~ install.sub > replace test command with [[ ]] > OK krw@ (rpe@) ~ install.sub > The data for CGI_{METHOD,TIME,TZ} comes from an external source. > Validate before assignment. > with feedback from and OK halex@ > OK krw@ (rpe@) ~ install.sub > Discard stderr of fdisk and pdisk. > disussed with and OK krw@ (rpe@) sets ~ lists/base/md.alpha ~ lists/base/md.amd64 ~ lists/base/md.armish ~ lists/base/md.armv7 ~ lists/base/md.hppa ~ lists/base/md.hppa64 ~ lists/base/md.i386 ~ lists/base/md.landisk ~ lists/base/md.loongson ~ lists/base/md.luna88k ~ lists/base/md.macppc ~ lists/base/md.octeon ~ lists/base/md.sgi ~ lists/base/md.socppc ~ lists/base/md.sparc ~ lists/base/md.sparc64 ~ lists/base/md.vax ~ lists/base/md.zaurus ~ lists/comp/mi > sync (deraadt@) ~ lists/base/mi > sync (deraadt@) ~ lists/man/mi > sync (deraadt@) special ~ pdisk/Makefile > get caught up with files krw removed (deraadt@) ~ pdisk/Makefile > Use normal err/errx/warn/warnx rather than handrolling new ones. Move > usage() to pdisk.c and nuke do_help() and '-h'. > ok jasper@ deraadt@ (krw@) ~ pdisk/Makefile > Nuke 'HFS_name' field which was not referenced now other than setting > it and freeing it. Unhook hfs_misc.c from build since it now longer > contains anything being used. (krw@) ~ pdisk/Makefile > Replace bitfield_get() and bitfield_set() abstraction with simple masking > operations since all the uses but 1 are for 1 bit fields. No functional > change intended. > Unhook bitfield.c from build. (krw@) ~ pdisk/Makefile > Nuke support for 1024- and 2048-byte sector devices. Error out if > anything other than 512-byte sectors are encountered. Unlink > deblock_media.c from the build. > ok jasper@ (krw@) ~ pdisk/Makefile > No need to abstract malloc(), so replace new_file_media() with > malloc(). > Removes last use of last used function in media.c, so unhook that > from the Makefile and remove forward declarations from media.h. > No functional change. (krw@) == games ============================================================= 03/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/games arithmetic ~ arithmetic.c > As per style(9), remove remaining lint-style comments from games/ > OK tb@ (mestre@) boggle ~ boggle/bog.c ~ boggle/help.c ~ boggle/mach.c ~ boggle/timer.c ~ boggle/word.c > Remove unnecessary headers > OK tb@ (mestre@) ~ boggle/mach.c ~ boggle/timer.c > As per style(9), remove remaining lint-style comments from games/ > OK tb@ (mestre@) ~ boggle/help.c ~ boggle/mach.c > Remove needless casts to (char *) NULL. Some were kept, namely for execl(3) > last arguments since "The list of arguments must be terminated by a null > null pointer" of type char * > Suggested by and OK tb@ (mestre@) bs ~ bs.c > Remove needless casts to (char *) NULL. Some were kept, namely for execl(3) > last arguments since "The list of arguments must be terminated by a null > null pointer" of type char * > Suggested by and OK tb@ (mestre@) canfield ~ canfield/canfield.c > As per style(9), remove remaining lint-style comments from games/ > OK tb@ (mestre@) cribbage ~ instr.c ~ io.c > As per style(9), remove remaining lint-style comments from games/ > OK tb@ (mestre@) fish ~ fish.c > As per style(9), remove remaining lint-style comments from games/ > OK tb@ (mestre@) fortune ~ fortune/fortune.c ~ unstr/unstr.c > As per style(9), remove remaining lint-style comments from games/ > OK tb@ (mestre@) hack ~ hack.c > #ifdef QUEST then a few functions won't work, because of that include > <stdlib.h> if we are on a QUEST > OK tb@ (mestre@) hangman ~ main.c > As per style(9), remove remaining lint-style comments from games/ > OK tb@ (mestre@) hunt ~ hunt/playit.c ~ huntd/answer.c > As per style(9), remove remaining lint-style comments from games/ > OK tb@ (mestre@) ~ hunt/hunt.c > Remove needless casts to (char *) NULL. Some were kept, namely for execl(3) > last arguments since "The list of arguments must be terminated by a null > null pointer" of type char * > Suggested by and OK tb@ (mestre@) phantasia ~ fight.c ~ io.c ~ main.c ~ misc.c ~ setup.c > As per style(9), remove remaining lint-style comments from games/ > OK tb@ (mestre@) quiz ~ quiz.c > As per style(9), remove remaining lint-style comments from games/ > OK tb@ (mestre@) random ~ random.c > As per style(9), remove remaining lint-style comments from games/ > OK tb@ (mestre@) sail ~ lo_main.c > there's a resource leak here. missing a fclose(3) to kill the stream. > OK deraadt@ mestre@ (gsoares@) tetris ~ input.c > As per style(9), remove remaining lint-style comments from games/ > OK tb@ (mestre@) wump ~ wump.c > As per style(9), remove remaining lint-style comments from games/ > OK tb@ (mestre@) == gnu =============================================================== 04/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/gnu gnu ~ gcc/gcc/config/mt/mt.c > fix typo > from Jan Schreiber, ok deraadt@ (mmcc@) usr.bin/perl ~ dist/PathTools/Cwd.pm ~ dist/PathTools/Cwd.xs ~ dist/PathTools/lib/File/Spec.pm ~ dist/PathTools/lib/File/Spec/Cygwin.pm ~ dist/PathTools/lib/File/Spec/Epoc.pm ~ dist/PathTools/lib/File/Spec/Functions.pm ~ dist/PathTools/lib/File/Spec/Mac.pm ~ dist/PathTools/lib/File/Spec/OS2.pm ~ dist/PathTools/lib/File/Spec/Unix.pm ~ dist/PathTools/lib/File/Spec/VMS.pm ~ dist/PathTools/lib/File/Spec/Win32.pm ~ dist/PathTools/t/taint.t > Fix perl CVE-2015-8607 - XS File::Spec::canonpath loses taint > https://rt.perl.org/Ticket/Display.html?id=126862 (afresh1@) ~ dist/PathTools/Cwd.pm ~ dist/PathTools/Cwd.xs ~ dist/PathTools/lib/File/Spec.pm ~ dist/PathTools/lib/File/Spec/Cygwin.pm ~ dist/PathTools/lib/File/Spec/Epoc.pm ~ dist/PathTools/lib/File/Spec/Functions.pm ~ dist/PathTools/lib/File/Spec/Mac.pm ~ dist/PathTools/lib/File/Spec/OS2.pm ~ dist/PathTools/lib/File/Spec/Unix.pm ~ dist/PathTools/lib/File/Spec/VMS.pm ~ dist/PathTools/lib/File/Spec/Win32.pm ~ dist/PathTools/t/taint.t TAGGED OPENBSD_5_8 > Fix perl CVE-2015-8607 - XS File::Spec::canonpath loses taint > https://rt.perl.org/Ticket/Display.html?id=126862 (afresh1@) ~ dist/PathTools/Cwd.pm ~ dist/PathTools/Cwd.xs ~ dist/PathTools/lib/File/Spec.pm ~ dist/PathTools/lib/File/Spec/Cygwin.pm ~ dist/PathTools/lib/File/Spec/Epoc.pm ~ dist/PathTools/lib/File/Spec/Functions.pm ~ dist/PathTools/lib/File/Spec/Mac.pm ~ dist/PathTools/lib/File/Spec/OS2.pm ~ dist/PathTools/lib/File/Spec/Unix.pm ~ dist/PathTools/lib/File/Spec/VMS.pm ~ dist/PathTools/lib/File/Spec/Win32.pm ~ dist/PathTools/t/taint.t TAGGED OPENBSD_5_7 > Fix perl CVE-2015-8607 - XS File::Spec::canonpath loses taint > https://rt.perl.org/Ticket/Display.html?id=126862 > Also minor changes from perl 5.20.2 (afresh1@) == regress =========================================================== 05/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/regress bin ~ ps/command.sh > some UTF-8 tests (schwarze@) ~ ksh/Makefile + ksh/vi/Makefile + ksh/vi/config.h + ksh/vi/edit.h + ksh/vi/sh.h + ksh/vi/test_vi.c + ksh/vi/vi.sh > The ksh(1) vi editing mode code is much harder to understand than > the emacs editing mode code, so add a testsuite before touching it. > To avoid having to deal with the horrendous low-level terminal > handling and high-level command execution code, write the tests in > unit test style rather than in integration test style, by providing > minimal glue to run vi.c stand-alone. In case relevant internal > interfaces in sh.c, edit.c, and friends are changed, the glue may > need adjustment. > Not testing completion and history at this time, only line editing. > (schwarze@) ~ ksh/vi/edit.h ~ ksh/vi/test_vi.c > better mockup for promptlen() (schwarze@) usr.bin ~ colrm/colrm.sh > Make the test file 7-bit clean and improve diagnostics with hexdump(1), > both in the same way as it was done for ul(1). > No functional change. (schwarze@) ~ colrm/colrm.sh > tests for double backspace after double-width characters (schwarze@) == sbin ============================================================== 06/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sbin ifconfig ~ ifconfig.8 ~ ifconfig.c > Make 'ifconfig $if mode' a valid subcommand that works independently of > the 'media' subcommand. Allow clearing the mode with 'ifconfig $if -mode'. > This makes commands such as 'ifconfig iwn0 mode 11a' work without having > to type all of 'ifconfig iwn0 media autoselect mode 11a'. > ok sthen@ deraadt@ jmc@ (stsp@) isakmpd ~ isakmpd.policy.5 > typo fix; from julian hsiao (jmc@) pdisk ~ partition_map.c > Use the normal method for determining a disk size. i.e. use the > device's default disklabel. Don't spew i/o's at the disk to find > the largest address that works, displaying many a 'read failed' > message along the way. > Problem noted by Brian Conway on bugs@ and jasper@. > ok jasper@ (krw@) ~ bitfield.c ~ bitfield.h ~ convert.c ~ convert.h ~ deblock_media.c ~ deblock_media.h ~ dpme.h ~ dump.c ~ dump.h ~ errors.c ~ errors.h ~ file_media.c ~ file_media.h ~ hfs_misc.c ~ hfs_misc.h ~ io.c ~ io.h ~ media.c ~ media.h ~ partition_map.c ~ partition_map.h ~ pathname.c ~ pathname.h ~ pdisk.c ~ pdisk.h ~ util.c ~ util.h ~ validate.c ~ validate.h ~ version.h > add rcs ids (jasper@) ~ file_media.c ~ io.c ~ partition_map.c ~ validate.c > trailing whitespace (jasper@) ~ pdisk.c > more whitespace was lurking (jasper@) ~ file_media.c > plug a mem leak > ok krw@ (jasper@) ~ dpme.h ~ file_media.c ~ hfs_misc.c > - remove some unused data structures/fields > - unifdef notdef > ok krw@ (jasper@) ~ Makefile ~ partition_map.c ~ pdisk.c ~ pdisk.8 > Remove '-v' and related version handling. Use strncasecmp() instead > of handrolled local istrncmp(). No need to compile util.c anymore. > ok deraadt@ jasper@ (krw@) - util.c - util.h - version.h > Nuke now unused files. > ok deraadt@ jasper@ (krw@) ~ Makefile ~ dump.c ~ partition_map.c ~ pdisk.c ~ validate.c > open_pathname_as_media() simply called open_file_as_media() which > takes the same parameters and returns the same value. Nuke > open_pathname_as_media() and call open_file_as_media() directly. > Nothing else from pathname.[ch] is used so remove them from build > in preparation for moving them to the attic. > ok deraadt@ (krw@) - pathname.c - pathname.h > Nuke now unused files. > ok deraadt@ (krw@) ~ pdisk.c > Remove some '-v' leftovers. Pointed out by jasper@. (krw@) ~ errors.c > More '-v' leftovers (usage). (krw@) ~ partition_map.c ~ pdisk.c > Since we use the disklabel size info we are always in 'device size > mode' and '-c' and the '-d' command to toggle cflag are pointless > since cflag is never examined. (krw@) ~ dump.c ~ errors.c ~ partition_map.c ~ partition_map.h ~ pdisk.8 ~ pdisk.c > Retire '-i', a poorly documented interactive mode that tries to > mimic the 20th century MacOS version of pdisk more closely than the > usual interactive editor. One set of commands is enough for anyone. > ok jasper@ (krw@) ~ file_media.c ~ file_media.h ~ media.c ~ media.h > Remove media interation types, structures and functions. They are not used > anywhere, as evidenced by everything still compiling after they are gone. > (krw@) ~ Makefile ~ dump.c ~ file_media.c ~ hfs_misc.c ~ io.c ~ partition_map.c ~ pdisk.8 ~ pdisk.c ~ validate.c > Use normal err/errx/warn/warnx rather than handrolling new ones. Move > usage() to pdisk.c and nuke do_help() and '-h'. > ok jasper@ deraadt@ (krw@) - errors.c - errors.h > Move unused files to attic. (krw@) ~ pdisk.c > The only thing in pdisk.h is an unused #define. So don't include it. (krw@) - pdisk.h > Another unused file for the attic. (krw@) ~ file_media.c ~ io.c > #define LOFF_T off_t > #define llseek lseek > #define LOFF_T_MAX LLONG_MAX > I mean, really? Nuke'em and use off_t, lseek and LLONG_MAX. > Also make a couple of checks for overflow of off_t value more likely > to work. > ok millert@ (krw@) ~ file_media.c > Nuke unused enum. (krw@) ~ pdisk.c > Remove obsolete flags (h, v, i, c) from getopt() call. > No need for 'flags' variable. Just call usage() on unknown options. (krw@) ~ dump.c ~ dump.h ~ pdisk.c > Remove code supporting undocumented and inaccessible flags 'a' > (abbreviate partition type names), 'f' (print HFS partition names) > and 'kLogicalOption' (obviously a getopt_long() conversion error). (krw@) ~ Makefile ~ partition_map.c ~ partition_map.h > Nuke 'HFS_name' field which was not referenced now other than setting > it and freeing it. Unhook hfs_misc.c from build since it now longer > contains anything being used. (krw@) ~ partition_map.c ~ pdisk.c > Don't include hfs_misc.h, which no longer contains anything of use. (krw@) - hfs_misc.c - hfs_misc.h > Move more unused files to attic. (krw@) ~ pdisk.c > Kill unused variable 'lfile'. (krw@) ~ Makefile ~ dpme.h ~ dump.c ~ partition_map.c > Replace bitfield_get() and bitfield_set() abstraction with simple masking > operations since all the uses but 1 are for 1 bit fields. No functional > change intended. > Unhook bitfield.c from build. (krw@) - bitfield.c - bitfield.h > Two more for the attic. (krw@) ~ dump.c ~ partition_map.c ~ partition_map.h ~ pdisk.c ~ validate.c > We don't need yet another alias for DEV_BSIZE. Use DEV_BSIZE and > nuke PBLOCK_SIZE and another unused #define. (krw@) ~ file_media.c > As with compute_device_size(), gut compute_block_size() and simply > ask the disklabel as everyone else does. Add file name to > compute_block_size() parameters to make error messages nicer. (krw@) ~ Makefile ~ file_media.c ~ partition_map.c ~ validate.c > Nuke support for 1024- and 2048-byte sector devices. Error out if > anything other than 512-byte sectors are encountered. Unlink > deblock_media.c from the build. > ok jasper@ (krw@) - deblock_media.c - deblock_media.h > Another couple for the attic. (krw@) - HISTORY - README > remove these files for they're no longer accurate, up to date or relevant > for > that matter. > ok krw@ (jasper@) ~ dump.c ~ file_media.c ~ file_media.h ~ media.c ~ media.h ~ partition_map.c ~ pdisk.c ~ validate.c > Start peeling away excess layers of abstraction. > Since 'file' is the only kind of media, no need to call the appropriate > functions via pointers. Just call the _file_ variants directly. > Nuke the fields do_read(), do_write(), do_close(), do_os_reload() > and the functions read_media(), write_media(), close_media() and > os_reload_media(). > No functional change. (krw@) ~ media.h > Remove some forward declarations missed in last commit. (krw@) ~ file_media.c ~ media.c ~ media.h > There can be only one. 'kind' that is. No need to allocate or track the > media kind since it will always be the same -- a file. > No functional change. (krw@) ~ file_media.c ~ media.c ~ media.h ~ partition_map.c ~ pdisk.c ~ validate.c > There can be only one. 'grain' that is. i.e. DEV_BSIZE. No need to track > or store. Just use. > No functional change. (krw@) ~ Makefile ~ media.h > No need to abstract malloc(), so replace new_file_media() with > malloc(). > Removes last use of last used function in media.c, so unhook that > from the Makefile and remove forward declarations from media.h. > No functional change. (krw@) - media.c > Another to the attic. (krw@) ~ file_media.c > Oops. Missed a file. (krw@) ~ dump.c ~ file_media.c ~ file_media.h ~ partition_map.c ~ partition_map.h ~ pdisk.c ~ validate.c > Move last field (size_in_bytes) in struct media to struct file_media > and replace MEDIA with FILE_MEDIA everywhere. media.h becomes unused. > (krw@) - media.h > Another one to the attic. (krw@) ~ Makefile ~ dump.c ~ partition_map.c > - tidy up a bunch of #if's for code that has been either enabled or > disabled since forever, > or not relevant at all (e.g. some bits for A/UX support) > - remove commented code that has never been compiled, etc > ok krw@ (jasper@) ~ Makefile > oops (jasper@) ~ convert.c ~ convert.h ~ dpme.h ~ dump.c ~ partition_map.c ~ partition_map.h > A/UX is no more, was m68k only, and we have no sentimental reasons for > keeping support. (krw@) ~ dump.c ~ file_media.c ~ file_media.h ~ partition_map.c ~ partition_map.h ~ pdisk.c ~ validate.c > Let the great de-typedef'ification begin. MEDIA -> struct file_media *. > (krw@) ~ convert.c ~ convert.h ~ dpme.h ~ dump.c ~ partition_map.c ~ partition_map.h ~ pdisk.c ~ validate.c > The great de-typedef'ification continues. DPME -> struct dpme. (krw@) ~ convert.c ~ convert.h ~ dpme.h ~ dump.c ~ partition_map.c ~ partition_map.h ~ pdisk.c ~ validate.c > The great de-typedef'ification continues. Block0 -> struct block0. (krw@) ~ convert.c ~ dpme.h ~ dump.c ~ partition_map.c > The great de-typedef'ification continues. DDMap -> struct ddmap. (krw@) ~ dump.c > The great de-typedef'ification continues. OSType -> unsigned long. > Nuke names/NAMES as the only use was in an unused array of abbreviations > which also goes. (krw@) ~ dump.c > The great de-typedef'ification continues. > PatchList -> struct patchlist > Nuke PatchListPtr > PatchListDescriptor -> struct patchlistdescriptor > Nuke PatchListDescriptorPtr (krw@) ~ validate.c > The great de-typedef'ification continues. range_list -> struct range_list. > (krw@) ~ dump.c ~ dump.h ~ partition_map.c ~ partition_map.h ~ pdisk.c ~ validate.c ~ validate.h > The great de-typedef'ification continues. > partition_map -> struct partition_map > partition_map_header -> struct partition_map_header (krw@) ~ convert.c ~ dpme.h > The great de-typedef'ification continues. u8 -> uint8_t. (krw@) ~ convert.c ~ dpme.h > The great de-typedef'ification continues. u16 -> uint16_t. (krw@) ~ dpme.h ~ dump.c ~ partition_map.c ~ partition_map.h ~ validate.c > The great de-typedef'ification concludes with u32 -> uint32_t. And a > bunch of consequent printf() format corrections. (krw@) ~ convert.c ~ convert.h ~ dpme.h ~ dump.c ~ dump.h ~ file_media.c ~ file_media.h ~ io.c ~ io.h ~ partition_map.c ~ partition_map.h ~ pdisk.c ~ validate.h ~ validate.c > Remove a bunch of ridiculous and uninformative comments like "// Defines". > (krw@) ~ dpme.h ~ dump.c ~ io.c ~ partition_map.c ~ pdisk.c ~ validate.c > Kill leading, imbedded, trailing whitespace. (krw@) ~ convert.c ~ convert.h ~ dump.c ~ dump.h ~ file_media.c ~ file_media.h ~ io.c ~ io.h ~ partition_map.c ~ partition_map.h ~ pdisk.c ~ validate.c ~ validate.h > Reduce namespace pollution by eliminating parameter names from forward > declarations. (krw@) ~ convert.c ~ convert.h ~ dpme.h ~ dump.c ~ dump.h ~ file_media.c ~ io.c ~ io.h ~ partition_map.c ~ partition_map.h ~ pdisk.c ~ validate.c ~ validate.h > Change "//" comments to "/**/". Nuke really stupid ones. (krw@) pfctl ~ pfctl.c > detect multiple root queues on a single interface and give a nice error > message - beats "pfctl: DIOCXCOMMIT: Invalid argument". > from Nathanael Rensen <nathanael at list.polymorpheus.com>, 10x! > ok sthen phessler, commit reminder mikeb (henning@) pflogd ~ pflogd.8 ~ pflogd.c ~ privsep.c > Interface status printing (at exit and USR1) was broken for a while. > Remove it completely to simplify the code. > even better deraadt@ (canacar@) == share ============================================================= 07/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/share man ~ man4/sdhc.4 > Mention sdhc* at acpi?. (kettenis@) ~ man4/ifmedia.4 > Remove wireless turbo mode support. It is a non-standard extension > which only worked with ath(4) devices from a decade ago. > Diff tested on ath(4) hardware by me to verify that 11a/b modes still work. > ok reyk deraadt chris sthen kettenis (stsp@) ~ man5/malloc.conf.5 > accurately describe that the (3) malloc option settings are only > inspected on the first call to a malloc(3) family function. > ok tedu (deraadt@) ~ man5/malloc.conf.5 > zap trailing whitespace (deraadt@) ~ man4/xnf.4 > Drop implementation notes section (mikeb@) ~ man4/ddb.4 > document "show socket". > jmc@ was super keen for this. (dlg@) ~ man4/xnf.4 > All emulated networking interfaces are taken over by xnf(4) now > This follows the best practice and was brought to our attention by > Wei Liu <wei ! liu2 at citrix ! com>, thanks! Implementation idea > from reyk@. (mikeb@) ~ man4/iwm.4 ~ man4/iwn.4 > add missing HISTORY section to iwm and iwn man pages; ok jmc (tj@) ~ man4/puc.4 > Moxa CP-168U support; from Olaf Schreck (deraadt@) ~ man4/xnf.4 > add missing article; (jmc@) ~ man4/Makefile + man4/dwiic.4 + man4/ihidev.4 + man4/ims.4 + man4/ikbd.4 > add pages for dwiic, ihidev, ims and ikbd (jsg@) ~ man4/iic.4 > mention dwiic and ihidev (jsg@) ~ man4/ikbd.4 > provide missing section number to Dt macro; (jmc@) mk ~ bsd.README > fix remaining sudo reference (s,sudo,doas,) > OK deraadt@ millert@ tb@ sthen@ (gsoares@) == sys =============================================================== 08/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sys arch/alpha/pci ~ tsvar.h > dev/isa/isavar.h does not appear to be needed by any of consumers. > (noticed because of another issue: one consumer is a high level busdriver, > but also contains i2c support, so the isa #define for ia_irq clashes > against i2c) (deraadt@) arch/amd64/amd64 ~ acpi_machdep.c > Introduce acpi_intr_establish(9), an API to establish interrupt handlers > for > "platform" devices that attach to acpi(4). > ok mlarkin@, deraadt@ (kettenis@) ~ vmm.c > Page fault handling tweaks for vmm: > - compute fault reason for uvm_fault() (e.g. page not present, > protection violation) instead of passing a protection code > - a page does not need to be zero'd after faulting it in. uvm_fault() does > that for fresh anon pages already, and we also do not want a page that > is swapped back in have its contents wiped. > ok mlarkin@ (stefan@) arch/amd64/conf ~ GENERIC > Add a "mux -1" locator to wsdisplay(4) instances representing secondary > displays. This makes sure the keyboard mux gets picked up by the primary > (console) display and that USB keyboards get paired with the console even > if they are not marked as the console keyboard. This is what most people > expect to happen. If you really want to associate your USB keyboard with > a secondary display, you can detach it from the mux using > wsconscfg -f /dev/ttyCcfg -k -d > and attach it to another display using > wsconscfg -f /dev/ttyDcfg -k (kettenis@) ~ GENERIC > Enable sdhc* at acpi?. (kettenis@) ~ GENERIC > Add dwiic, a driver for the Synopsys DesignWare i2c controller found > on the Samsung ATIV Book 9 laptop. This initial version only > supports ACPI config/attachment. > Add ihidev, a HID-over-i2c driver largely based on uhidev. dwiic > handles attaching ihidev devices found in ACPI. > Add ims, a HID-over-i2c mouse/trackpad driver to get basic cursor > and button functionality from HID-compliant i2c trackpads. > ok kettenis deraadt (jcs@) ~ GENERIC > Enable Xen support in GENERIC; discussed with deraadt@ (mikeb@) ~ GENERIC > Eable ikbd(4). (kettenis@) ~ RAMDISK_CD > enable sdhc@acpi so emmc will be available at install time > ok kettenis@ (jsg@) ~ RAMDISK_CD > enable ikbd(4) (jsg@) arch/amd64/include ~ specialreg.h > Page fault handling tweaks for vmm: > - compute fault reason for uvm_fault() (e.g. page not present, > protection violation) instead of passing a protection code > - a page does not need to be zero'd after faulting it in. uvm_fault() does > that for fresh anon pages already, and we also do not want a page that > is swapped back in have its contents wiped. > ok mlarkin@ (stefan@) arch/amd64/stand/libsa ~ random_amd64.S ~ random_i386.S > match up code and comments, add missing $OpenBSD$; ok jsing@ a while ago > (naddy@) ~ random_amd64.S ~ random_i386.S > naddy observed the TSC fallback code could encounter (high word) %edx > being 0; after multiply there is no perturbance to the from-disk entropy > buffer. Further investigation showed perturbance was biased towards > the lower bytes of a word. Compensate for this with a hocus pocus > bit-spreading operation which applies a result byte by byte. > discussion with kettenis, tb, mlarkin, naddy > ok naddy (deraadt@) arch/armish/conf ~ files.armish > Move files.hid above files.i2c to fix armish build (tobiasu@) arch/armv7/exynos ~ exesdhc.c > Seems the voltage bit defenitions in sdmmcreg.h for voltages below 2V > are not quite right. At least I can't find them in any of the MMC and > SD card documentation I can find on the interwebs. Instead there is a > single "low voltage bit" that indicates support for the 1.65-1.95V or > 1.70-1.95V range depending on the document you're reading. Go with the > 1.65-1.95V range as that is what Linux does. > Necessary (but not sufficient) to make the eMMC on the ASUS X205TA work. > ok jsg@ (who did the armv7 bits) (kettenis@) arch/armv7/imx ~ imxesdhc.c > Seems the voltage bit defenitions in sdmmcreg.h for voltages below 2V > are not quite right. At least I can't find them in any of the MMC and > SD card documentation I can find on the interwebs. Instead there is a > single "low voltage bit" that indicates support for the 1.65-1.95V or > 1.70-1.95V range depending on the document you're reading. Go with the > 1.65-1.95V range as that is what Linux does. > Necessary (but not sufficient) to make the eMMC on the ASUS X205TA work. > ok jsg@ (who did the armv7 bits) (kettenis@) arch/armv7/omap ~ ommmc.c > Seems the voltage bit defenitions in sdmmcreg.h for voltages below 2V > are not quite right. At least I can't find them in any of the MMC and > SD card documentation I can find on the interwebs. Instead there is a > single "low voltage bit" that indicates support for the 1.65-1.95V or > 1.70-1.95V range depending on the document you're reading. Go with the > 1.65-1.95V range as that is what Linux does. > Necessary (but not sufficient) to make the eMMC on the ASUS X205TA work. > ok jsg@ (who did the armv7 bits) (kettenis@) arch/i386/conf ~ GENERIC > Add a "mux -1" locator to wsdisplay(4) instances representing secondary > displays. This makes sure the keyboard mux gets picked up by the primary > (console) display and that USB keyboards get paired with the console even > if they are not marked as the console keyboard. This is what most people > expect to happen. If you really want to associate your USB keyboard with > a secondary display, you can detach it from the mux using > wsconscfg -f /dev/ttyCcfg -k -d > and attach it to another display using > wsconscfg -f /dev/ttyDcfg -k (kettenis@) arch/i386/stand/libsa ~ random_i386.S > match up code and comments, add missing $OpenBSD$; ok jsing@ a while ago > (naddy@) ~ random_i386.S > naddy observed the TSC fallback code could encounter (high word) %edx > being 0; after multiply there is no perturbance to the from-disk entropy > buffer. Further investigation showed perturbance was biased towards > the lower bytes of a word. Compensate for this with a hocus pocus > bit-spreading operation which applies a result byte by byte. > discussion with kettenis, tb, mlarkin, naddy > ok naddy (deraadt@) arch/loongson/conf ~ files.loongson > include files.hid before files.i2c (deraadt@) arch/mips64/include ~ pmap.h > Back out the MP pmap diff for rework. The code does not work on > non-PMAP_DIRECT systems due to lock recursion. (visa@) arch/mips64/mips64 ~ pmap.c ~ trap.c > Back out the MP pmap diff for rework. The code does not work on > non-PMAP_DIRECT systems due to lock recursion. (visa@) arch/octeon/conf ~ GENERIC ~ RAMDISK ~ files.octeon > Add support for a variety of USB devices. > Initial diff from Artturi Alm, ok jasper@ (visa@) arch/octeon/dev ~ mainbus.c > Attach secondary CPUs by coremask. This fixes a crash that happened if > the MP kernel was booted with coremask=1. (visa@) arch/octeon/octeon ~ machdep.c > Panic if someone boots the kernel by force without CPU 0. The system > will not work without that core. (visa@) arch/sh/include ~ intr.h > intr_barrier for sh/landisk; tobiasu noticed the need as well (deraadt@) arch/sh/sh ~ interrupt.c > intr_barrier for sh/landisk; tobiasu noticed the need as well (deraadt@) ddb ~ db_command.c ~ db_interface.h > add a "show socket" command to ddb > should help inspecting socket issues in the future. > enthusiasm from mpi@ bluhm@ deraadt@ (dlg@) dev ~ ipmi.c ~ ipmivar.h > When stopping wdog (via wdog_shutdown() e.g. before entering shutdown), > clear DONTSTOP bits. Without this, BMC records, at least on some NEC > servers, watchdog timer expiration event, which only confuses admins. > OK kettenis@ (uebayasi@) ~ ipmi.c > Print a message to console when ipmi(4) watchdog is either enabled or > disabled. Replace a magic number while here. (uebayasi@) ~ ipmi.c > Read values from disabled sensors if possible. From jmatthew@. (uebayasi@) ~ ipmi.c ~ ipmivar.h > Remove useless sc_poll flag because it is always 1. Remove unused global > ipmi_poll too. (uebayasi@) ~ ipmi.c ~ ipmivar.h > Introduce struct ipmi_cmd and have ipmi_cmd() function. IPMI command > sequence is always a pair of send and receive; call both from ipmi_cmd(). > By making ipmi_cmd() take only one argument struct ipmi_cmd *, this also > helps to prepare to call ipmi_cmd() as a task. > No functional changes intended. (uebayasi@) ~ ipmi.c ~ ipmivar.h > Refactor buildmsg() functions to take struct ipmi_cmd * instead of 6 > arguments. No functional changes. (uebayasi@) ~ ipmi.c ~ ipmivar.h > Allocate command data buffer on softc and use it from both ipmi_sendcmd() > and ipmi_recvcmd(). No functional changes. (uebayasi@) ~ ipmi.c ~ ipmivar.h > Make sendmsg() and recvmsg() functions take only struct ipmi_cmd *. > Handle interface specific command length more nicely. No functional > changes. (uebayasi@) ~ ksyms.c > Use uiomove() instead of uiomovei(); from Martin Natano > ok kettenis@ (stefan@) ~ ipmi.c > Refactor ipmi_watchdog(); split tickle and set into separate functions. > (uebayasi@) ~ ipmi.c ~ ipmivar.h > Use task to execute watchdog tickle. (uebayasi@) ~ ipmi.c ~ ipmivar.h > Use task to execute command except polling context. (uebayasi@) dev/acpi ~ acpithinkpad.c > Add thinkpad tablet dock/undock ACPI events. > ok deraadt kettenis (stsp@) ~ acpivar.h > Introduce acpi_intr_establish(9), an API to establish interrupt handlers > for > "platform" devices that attach to acpi(4). > ok mlarkin@, deraadt@ (kettenis@) ~ files.acpi + sdhc_acpi.c > Add glue to make sdhc(4) attach to acpi(4). Needed to support the SD host > controllers integrated on Intel's Bay Trail SoCs. (kettenis@) ~ files.acpi > Only include sdhc_acpi.c if shdc* at acpi? is included in the kernel > config. (kettenis@) ~ acpi.c > Make sdhc(4) attach to hardware IDs "80860F14" and "PNP0FFF". (kettenis@) ~ acpi.c ~ acpireg.h ~ acpivar.h ~ files.acpi + dwiic.c > Add dwiic, a driver for the Synopsys DesignWare i2c controller found > on the Samsung ATIV Book 9 laptop. This initial version only > supports ACPI config/attachment. > Add ihidev, a HID-over-i2c driver largely based on uhidev. dwiic > handles attaching ihidev devices found in ACPI. > Add ims, a HID-over-i2c mouse/trackpad driver to get basic cursor > and button functionality from HID-compliant i2c trackpads. > ok kettenis deraadt (jcs@) ~ acpi.c ~ acpireg.h > Improve emulation of PM registers on Hardware-redcued ACPI by emulating > SPL_TYPx, SLP_EN and WAK_STS using the new SLEEP_STATUS_REG and > SLEEP_CONTROL_REG registers. Makes the transition into S5 (powerdown) work > on the ASUS X205TA. > ok deraadt@, mlarkin@ (kettenis@) ~ dwiic.c > rename new i2c_attach_args fields to avoid conflict with cpp magic in > isavar.h (a few parts of the tree include both) > ok jcs (deraadt@) ~ dsdt.h > Add data structures and defines for Generic and I2C Serial Bus Connection > descriptors. > ok jcs@ (kettenis@) ~ dwiic.c > Don't confuse the HID descriptor address (which really is an office into > the > register space of the i2c device) with the i2c address. For i2c busses > enumerated by ACPI we get the address from the I2C Serial Bus Connection > Resource Descriptor returned by the _CRS methide of the i2c device. Pass > the > HID descriptor address in the ia_size member. > ok jcs@ (kettenis@) ~ dsdt.c ~ dsdt.h > Change aml_find_node() such that it only walks down the tree and doesn't > traverse sideways. This seems to be what all callersexpect it to do, and > fixes a bug in dwiic(4) where it would try to access i2c devices on busses > they're not attached to. > If there is any fallout from this change, the right thing to do is probably > to make sure callers pass the right node. > While there, change the return type to void, as the return value was > useless > and none of the callers looked at it. > ok mlarkin@ (kettenis@) ~ dwiic.c > Several fixes for dwiic(4). > * Properly map bus space; using BUS_SPACE_MAP_PREFATCHABLE is not agood > idea > as it may lead to reordering or merging of register writes in the store > buffer. > * Properly implement the iic(4) operations in dwiic_i2c_exec(). > * Keep timings set up by the firmware if the SSCN and FMCN methods aren't > available. > ok jcs@ (kettenis@) ~ acpi.c ~ acpireg.h > Attach dwiic(4) on the Intel Bay Trail i2c controllers as well. (kettenis@) ~ dwiic.c > Check _STA method to see whether te device is actually present. If not, > make > the match function return 0 to prevent the driver from attaching. > (kettenis@) ~ dwiic.c > In dwiic_acpi_foundhid() treat _STA missing as if all the bits > were set as per the acpi spec. > ok kettenis@ (jsg@) ~ acpi.c > remove an unused global variable (jsg@) ~ acpi.c > Take acpi_foundhid() outside of #ifndef SMALL_KERNEL so the ramdisk > kernel can find and attach dwiic(4) and sdhc(4). (jsg@) ~ dwiic.c > dwiic_i2c_{read,write} -> dwiic_{read,write} (jcs@) ~ dwiic.c > fix an off-by-one that could result in read operations not reading > the last byte, depending on the timing (jcs@) dev/i2c ~ files.i2c ~ i2cvar.h + ihidev.c + ihidev.h + ims.c > Add dwiic, a driver for the Synopsys DesignWare i2c controller found > on the Samsung ATIV Book 9 laptop. This initial version only > supports ACPI config/attachment. > Add ihidev, a HID-over-i2c driver largely based on uhidev. dwiic > handles attaching ihidev devices found in ACPI. > Add ims, a HID-over-i2c mouse/trackpad driver to get basic cursor > and button functionality from HID-compliant i2c trackpads. > ok kettenis deraadt (jcs@) ~ i2cvar.h ~ ihidev.c > rename new i2c_attach_args fields to avoid conflict with cpp magic in > isavar.h (a few parts of the tree include both) > ok jcs (deraadt@) ~ ihidev.c ~ ihidev.h > Don't confuse the HID descriptor address (which really is an office into > the > register space of the i2c device) with the i2c address. For i2c busses > enumerated by ACPI we get the address from the I2C Serial Bus Connection > Resource Descriptor returned by the _CRS methide of the i2c device. Pass > the > HID descriptor address in the ia_size member. > ok jcs@ (kettenis@) ~ ihidev.c > fix hid packet length calculation > noticed by jsg (jcs@) ~ files.i2c + ikbd.c > Driver for HID-over-i2c keyboards. (kettenis@) ~ ihidev.c > When iterating over the report IDs to calculate the maximum size, take > into account the 2 bytes needed to store the length. Also make sure > that we have at least wMaxInputLength bytes available for reading input > reports. > ok jcs@ (kettenis@) ~ ihidev.c > Use I2C_F_POLL flag when acquiring and releasing the i2c bus in the > interrupt handler. > ok jcs@ (kettenis@) ~ ihidev.c > Several fixes for dwiic(4). > * Properly map bus space; using BUS_SPACE_MAP_PREFATCHABLE is not agood > idea > as it may lead to reordering or merging of register writes in the store > buffer. > * Properly implement the iic(4) operations in dwiic_i2c_exec(). > * Keep timings set up by the firmware if the SSCN and FMCN methods aren't > available. > ok jcs@ (kettenis@) dev/ic ~ ar5210.c ~ ar5210var.h ~ ar5211.c ~ ar5212.c ~ ar5xxx.c ~ ar5xxx.h ~ ath.c ~ pgt.c > Remove wireless turbo mode support. It is a non-standard extension > which only worked with ath(4) devices from a decade ago. > Diff tested on ath(4) hardware by me to verify that 11a/b modes still work. > ok reyk deraadt chris sthen kettenis (stsp@) ~ nvme.c > wrap up dma syncs for the whole mapping. (dlg@) ~ nvme.c ~ nvmereg.h > handle the version register like ahci. (dlg@) ~ nvmereg.h > fix calculation of the max queue entries supported value (dlg@) ~ nvme.c > feng shui. > dont need a billion tabs for the function prototypes. just some > akward line wrappings. (dlg@) ~ nvme.c > when enabling the controller, wait till CSTS.RDY lights up. (dlg@) dev/pci ~ if_em.c ~ if_em.h > do further work on the em transmit path to simplify the code. > noone could understand how em_txeof worked, so i rewrote it. > this also gets rid of the sc_tx_desc_free var that needed atomic > ops. space to use in em_start and space to free in em_txeof is now > calculated from the producer and consumer. > testers have reported better responsiveness with this. somehow. > if em issues persist after this, im rolling back to pre-mpsafe changes. > (dlg@) ~ if_em.c > post the packet on em_82547 chips after bpf > now that start and txeof can run on different cpus, txeof could > have freed the mbuf before bpf got to it. (dlg@) ~ if_iwn.c > fix a misplaced curly brace (stsp@) ~ if_iwn.c > Fix iwn(4) firmware SYSASSERT during hardware initialization. > For some reason we would end up passing channel number zero to > the RXON firmware command, which the firmware doesn't like. > ok sthen@ reyk@ > and also tested by stefan@, deraadt@, mikeb@ (stsp@) ~ if_iwn.c > Tell the iwn(4) firmware to retry failed Tx at 1Mbit/s instead of MCS 0. > Seems to make Tx rate scaling go up faster and helps Rx performance. > ok deraadt@ (stsp@) ~ if_iwn.c > Revert most of if_iwm.c r1.155. Use of channel zero was a symptom of > another bug introduced in my turbo mode removal commit; Now fixed. > Keep the debug printf added in r1.155. (stsp@) ~ eso.c > add brackets to avoid an operator precedence bug > ok ratchov@ deraadt@ (jsg@) ~ pcidevs > there are two types of samsung sm951 (dlg@) ~ pcidevs.h ~ pcidevs_data.h ~ ahci_pci.c > regen > ahci_pci updated to cope with the new pcidevs id for its sm951 (dlg@) ~ xspd.c > Detach emulated network devices if Netfront driver is enabled > Xen doesn't provide a way for a guest to decide which model of > the interface is selected in the VM configuration and therefore > there's no simple way for Netfront and emulated devices to co- > exist on the same system. Emulated em(4) or re(4) drivers will > take over if xnf(4) driver is disabled or not compiled in. > Idea and OK reyk (mikeb@) ~ pcidevs > Moxa CP-168U, from Olaf Schreck (deraadt@) ~ pcidevs.h ~ pcidevs_data.h > sync (deraadt@) ~ pucdata.c > Moxa CP-168U support; from Olaf Schreck (deraadt@) dev/pv ~ xenstore.c > Handle zero lenght messages in the xs_parse by returning an empty string > Problem was reported and analyzed by reyk@ (mikeb@) ~ xenstore.c > Do not fail when receiving an empty reply (or directory node) in > xs_intr() but put an empty message in the queue. This prevents > xs_reply() from being stuck in an endless loop because it expectes a > message in the queue to break out of it. Depends on mikeb@'s previous > commit because it would otherwise panic on trying to cleanup the empty > message. > OK mikeb@ (reyk@) ~ xenstore.c > xs_cmd should always return errno(2) codes (mikeb@) ~ xenstore.c > Convert XenStore code to free running producer/consumer indices > After some hair pulling while implementing xnf(4) I've realised that > Xen uses free running producer/consumer indices wrapping with their > type (unsigned 32 bit integer). Later I've confirmed it with free > chapters of the "The Definitive Guide to the Xen Hypervisor" by David > Chisnall available online. (mikeb@) ~ xenstore.c > A couple of unused variables have sneaked in from the debug code (mikeb@) ~ if_xnf.c ~ xen.c > Create rx and tx fragment maps with a page size boundary restriction > We need to ensure that rx and tx fragments do not cross page boundary > since grant table reference can only point to a complete page. Add a > couple of kernel assertions in the dma map loading code to catch these > problems early in the future. (mikeb@) ~ xen.c > spacing typos (mikeb@) ~ if_xnf.c > Implement transmit watchdog for testing purposes (mikeb@) ~ if_xnf.c > Bump number of tx fragments to the stock value of 18 (mikeb@) ~ xen.c > Fixup a merge error (mikeb@) ~ if_xnf.c ~ xen.c ~ xenvar.h > Detach emulated network devices if Netfront driver is enabled > Xen doesn't provide a way for a guest to decide which model of > the interface is selected in the VM configuration and therefore > there's no simple way for Netfront and emulated devices to co- > exist on the same system. Emulated em(4) or re(4) drivers will > take over if xnf(4) driver is disabled or not compiled in. > Idea and OK reyk (mikeb@) ~ xen.c ~ xenstore.c ~ xenvar.h > Cleanup dmesg output, disable debugging; prodding and suggestions from > reyk@ (mikeb@) dev/sdmmc ~ sdhc.c ~ sdmmcreg.h > Seems the voltage bit defenitions in sdmmcreg.h for voltages below 2V > are not quite right. At least I can't find them in any of the MMC and > SD card documentation I can find on the interwebs. Instead there is a > single "low voltage bit" that indicates support for the 1.65-1.95V or > 1.70-1.95V range depending on the document you're reading. Go with the > 1.65-1.95V range as that is what Linux does. > Necessary (but not sufficient) to make the eMMC on the ASUS X205TA work. > ok jsg@ (who did the armv7 bits) (kettenis@) ~ sdhc.c ~ sdhcreg.h > Add (limited) support for controllers that implement version 3.0 of the > SD host controller standard. Support the larger base clock and larger > clock divisors. > ok jsg@ (kettenis@) ~ sdmmc_cis.c ~ sdmmc_ioreg.h > Improve CIS parsing to recognize CISTPL_NULL codes instead of bailing out > on > them. Add symbolic constant for CISTPL_END. > ok jsg@ (kettenis@) dev/usb ~ ums.c ~ ukbd.c > properly initialize quirks before sending them to hid layer, broken > during hid abstraction > problem noticed and fix tested by abieber (jcs@) ~ usb_subr.c > Prevent a NULL dereference when detaching a USB device with ugen(4) > disabled or if allocating memory during the attachment process failed. > Problem reported by and original diff from James Hastings on bugs@. > ok deraadt@ (mpi@) dev/wscons ~ wsevent.c > Use uiomove() instead of uiomovei(); from Martin Natano > ok deraadt@ (stefan@) kern ~ subr_log.c > Convert to uiomove(); from Martin Natano, thanks! > ok deraadt@ (stefan@) ~ uipc_socket.c > To make bug hunting easier, print more information in the soreceive() > and somove() panic messages. > OK phessler@ benno@ deraadt@ mpi@ (bluhm@) ~ tty.c ~ tty_pty.c > Increase buffer sizes and watermarks for tty and ppp > Use 115200 the default speed for buffer sizing in ttymalloc(). A lot > of devices call ttymalloc(0) so this affects quite a few of them. > Increases the buffer size for 9600 < baud <= 115200 from 1k to 4k. > Make ppp use the lo/hi watermarks from the tty layer which are > adjusted according to speed + buffer size. The previous fixed values > of 100 and 400 were way too small > Make pty call ttymalloc() with baud == 1000000, which is the common > value used in the tree for "fast". > ok deraadt@ (sf@) ~ subr_pool.c ~ uipc_socket.c > add a "show socket" command to ddb > should help inspecting socket issues in the future. > enthusiasm from mpi@ bluhm@ deraadt@ (dlg@) ~ uipc_socket.c > print TAILQ_NEXT(so, so_qe) too (dlg@) ~ kern_synch.c > KASSERT on refcnt underflow. > ok mpi@ bluhm@ (dlg@) ~ uipc_socket.c > Improve the socket panic messages further. claudio@ wants to see > the socket type and dlg@ is interested in the pointers for ddb show > socket. > OK deraadt@ dlg@ (bluhm@) ~ sys_pipe.c > Convert to uiomove(); from Martin Natano > ok millert@ (stefan@) msdosfs ~ denode.h ~ msdosfs_vnops.c > Implement VFS read clustering for MSDOSFS. > The logic used in msdosfs_bmap() to loop calling pcbmap() comes from > FreeBSD and is not really efficient but it is good enough since it is > only called when generating I/O. > With this diff I get a 100% improvement when reading big files from a > crappy USB stick. > With this and bread_cluster(9) modified to not re-fetch B_CACHED buffers, > reading large contiguous files with chunk sizes of MAXPHYS is almost as > fast as physio(9) on the same device. > For a 'real world' example, when copying music files from a USB stick I > see a speed jump from 15MB/s on -current to 24Mb/s with this diff. > While here rename some 'lbn' variables into 'cn' to better reflect what > we're dealing with. > Tested by Mathieu, with support from deraadt@ (mpi@) net ~ rtsock.c > Compare if the prefix lengths are equal rather than checking if the > netmasks are non-NULL for RTM_CHANGE and RTM_LOCK. > Fix route(8) regressions tests 14 and 16 with ART kernels. These tests > were failing because with ART, RTF_HOST routes always have a mask. > ok claudio@, bluhm@ (mpi@) ~ if_media.h > Remove wireless turbo mode support. It is a non-standard extension > which only worked with ath(4) devices from a decade ago. > Diff tested on ath(4) hardware by me to verify that 11a/b modes still work. > ok reyk deraadt chris sthen kettenis (stsp@) ~ if_vlan.c > dont propagate any of the parent interfaces offload features on svlan > this avoids confusion about checksum offload if you configure ips > on svlan interfaces on top of interfaces that do offload, like bge > or em. > the problem was reported by and this fix was tested by denis fondras > analyzed by naddy@ > thanks guys (dlg@) ~ if_gif.c ~ if_mpe.c ~ if_pppx.c ~ if_vxlan.c > No need for a splnet() dance around IFQ_DEQUEUE() anymore. > From David Hill, ok dlg@ (mpi@) ~ ppp_tty.c > Increase buffer sizes and watermarks for tty and ppp > Use 115200 the default speed for buffer sizing in ttymalloc(). A lot > of devices call ttymalloc(0) so this affects quite a few of them. > Increases the buffer size for 9600 < baud <= 115200 from 1k to 4k. > Make ppp use the lo/hi watermarks from the tty layer which are > adjusted according to speed + buffer size. The previous fixed values > of 100 and 400 were way too small > Make pty call ttymalloc() with baud == 1000000, which is the common > value used in the tree for "fast". > ok deraadt@ (sf@) ~ pf_ruleset.c > rs_malloc() does not need to be cast (mmcc@) net80211 ~ ieee80211.c ~ ieee80211_ioctl.h ~ ieee80211_output.c ~ ieee80211_proto.c ~ ieee80211_radiotap.h ~ ieee80211_regdomain.h ~ ieee80211_var.h > Remove wireless turbo mode support. It is a non-standard extension > which only worked with ath(4) devices from a decade ago. > Diff tested on ath(4) hardware by me to verify that 11a/b modes still work. > ok reyk deraadt chris sthen kettenis (stsp@) ~ ieee80211.c > Fix fallout from turbo mode removal commit: Channel initialization > was broken and lots of drivers ran into trouble as a result. > ok deraadt@ (stsp@) netinet ~ ip_carp.c > Prevent a NULL-reference if SIOCGVH is issued without carpdev argument. > Problem reported by and diff tested by Fabian Raetz, thanks! > ok benno@, bluhm@ (mpi@) ~ ip_output.c > Prevent a double if_put(). > ok mikeb@, bluhm@ (mpi@) ~ if_ether.c > Grab the KERNEL_LOCK before delivering a message to the routing socket > when an ARP resolution has been done. > Should hopefully fix the "receive 1" panic reported by benno@ on bugs@. > ok claudio@, phessler@, benno@, bluhm@ (mpi@) netinet6 ~ nd6_rtr.c > Prevent a possible NULL-derefenced introduced by the move of RA sleeping > code in a task. > Such dereferenced can be triggered by receiving a RA with the 'on-link' > bit set to 0 apparently generated by dnsmasq and reported by matthieu@. > ok matthieu@, sthen@, bluhm@ (mpi@) ~ ip6_output.c > Prevent a double if_put(). > ok mikeb@, bluhm@ (mpi@) tmpfs ~ tmpfs_vfsops.c > kill malloc(9) pointless check, malloc(9) call is using M_WAITOK flag and > thus will sleep until > get memory available so shouln't fail. > OK deraadt@ mmcc@ (gsoares@) ufs/ext2fs ~ ext2fs_readwrite.c > Kill the "doclusterread" and "doclusterwrite" defines. > They are defined to 0 since the import of ext2fs 19 years ago. (mpi@) ufs/ffs ~ ffs_extern.h ~ ffs_vfsops.c ~ ffs_vnops.c > Remove the "doclusterread" and "doclusterwrite" buttons. > Cluster read is unconditionnally enabled for FFS since some years and > toggling "doclusterread" has no effect. > OpenBSD no longer support write clustering since the Dynamic Buffer > Cache went in, so there is reason to delay the call to bawrite(9). > Tested by Mathieu, ok dlg@ (mpi@) == usr.bin =========================================================== 09/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.bin aucat ~ abuf.c ~ afile.c ~ aucat.c ~ dsp.c ~ utils.c > trailing spaces, fit in 80 columns (ratchov@) calendar ~ calendars/calendar.music > :( David Bowie (sthen@) ~ calendar.1 ~ calendars/calendar.all ~ calendars/calendar.holiday + calendars/calendar.uk > add a calendar file for the uk; > this is the work of craig skinner - many thanks, craig. > i've tweaked it a bit. craig has also fixed a currently > incorrect date and moved a couple entries to the uk file/ > some outstanding issues (what to do with dups) but it'll be > eaiser to work on once in tree (jmc@) jot ~ Makefile ~ jot.c > Random sequences generated by jot -r have been suffering from a dramatic > bias at both ends of the given interval. Various clever workarounds are > documented in the manual. > This fixes the bias in some of the most common use cases, namely random > sequences of integers or real numbers rounded to a fixed number of > decimals. So, `jot -p 3 -r 10000 0 1' produces a uniformly distributed > sequence of 10,000 decimal numbers of the form 0.abc (or 1). > Internally, jot -r now uses arc4random_uniform() whenever this is > clearly possible. In particular `jot -r 1 10 20' yields an unbiased > random number between 10 and 20 (both ends inclusive) from the shell. > guidance and ok deraadt@, "i like it" benno@ (tb@) kdump ~ kdump_subr.h ~ ktrstruct.c ~ mksubr > Display struct event more like other structs: show all members and use cpp > define names. Use the mksubr infrastructure for EVFILT_*, > EV_*, and (with some kludging) NOTE_* values. If EV_ERROR is set, > include the errno string for the data member. > Also, do MSG_* parsing on the msg_flags member of struct msghdr. > ok millert@ (guenther@) less ~ cmdbuf.c ~ command.c ~ funcs.h ~ jump.c ~ less.h ~ line.c ~ linenum.c ~ output.c ~ prompt.c ~ search.c ~ tags.c > Remove LINUNUM, a needless alias for off_t. Also, remove a redundant > condition. > ok nicm@ (mmcc@) ~ edit.c ~ main.c ~ optfunc.c ~ prompt.c ~ signal.c > remove lint annotations > ok nicm@ (mmcc@) ~ less.h ~ line.c ~ output.c ~ prompt.c > unify two identical function pairs now that we've removed less's off_t > aliases > ok nicm@ (mmcc@) ~ less.h > Use quotes rather than angle brackets for local header. Fixes building > with Clang 3.7. > ok millert@ (mmcc@) ~ funcs.h > remove needless extern keywords, a vestige of pre-ANSI C > explained by and ok millert@ (mmcc@) ~ prompt.c > start removing mp, the pointer to message's nul byte > ok deraadt@ (mmcc@) mandoc ~ main.c > Unbreak reading from stdin after recent parse() restructuring. > OK schwarze@ (florian@) newsyslog ~ newsyslog.c > ajacoutot@ reported seeing > newsyslog: preposterous process number in pid file: /var/run/syslog.pid > in his log since syslogd ran as pid 2. > Nowadays only init runs with a reserved low pid, so decrease MIN_PID > from 4 to 2, as suggested by espie@. > ok kettenis@ (tb@) rpcgen ~ rpc_parse.c ~ rpc_scan.c ~ rpc_scan.h > Add support for parsing 'hyper' and 'quad' types, as per RFC4506. > From FreeBSD > ok millert@ (jasper@) ssh ~ sshbuf-getput-crypto.c ~ sshbuf.c > use explicit_bzero() more liberally in the buffer code; ok deraadt (djm@) ~ clientloop.c ~ clientloop.h ~ mux.c ~ ssh.c > eliminate fallback from untrusted X11 forwarding to trusted > forwarding when the X server disables the SECURITY extension; > Reported by Thomas Hoger; ok deraadt@ (djm@) ~ clientloop.c ~ clientloop.h ~ mux.c ~ ssh.c TAGGED OPENBSD_5_8 > MFC changes by [email protected], 2016/01/13 16:04:47 > "eliminate fallback from untrusted X11 forwarding to trusted > forwarding when the X server disables the SECURITY extension; > Reported by Thomas Hoger; ok deraadt@" (sthen@) ~ clientloop.c ~ clientloop.h ~ mux.c ~ ssh.c TAGGED OPENBSD_5_7 > MFC changes by [email protected], 2016/01/13 16:04:47 > "eliminate fallback from untrusted X11 forwarding to trusted > forwarding when the X server disables the SECURITY extension; > Reported by Thomas Hoger; ok deraadt@" (sthen@) ~ readconf.c ~ ssh.c TAGGED OPENBSD_5_7 > Disable experimental client-side roaming support. Server side was > disabled/gutted for years already, but this aspect was surprisingly > forgotten. > Thanks for report from Qualys (deraadt@) ~ readconf.c ~ ssh.c TAGGED OPENBSD_5_8 > MFC changes by: [email protected] 2016/01/14 07:34:34 > Disable experimental client-side roaming support. Server side was > disabled/gutted for years already, but this aspect was surprisingly > forgotten. > Thanks for report from Qualys (sthen@) ~ readconf.c ~ ssh.c TAGGED OPENBSD_5_7 > MFC changes by: [email protected] 2016/01/14 07:34:34 > Disable experimental client-side roaming support. Server side was > disabled/gutted for years already, but this aspect was surprisingly > forgotten. > Thanks for report from Qualys (sthen@) - roaming.h - roaming_client.c - roaming_common.c - roaming_dummy.c - roaming_serv.c ~ clientloop.c ~ kex.c ~ kex.h ~ monitor.c ~ monitor_wrap.c ~ opacket.c ~ opacket.h ~ packet.c ~ packet.h ~ readconf.c ~ readconf.h ~ serverloop.c ~ ssh.c ~ ssh2.h ~ sshconnect.c ~ sshconnect2.c ~ sshd.c ~ lib/Makefile ~ ssh/Makefile ~ ssh-keyscan/Makefile ~ ssh-keysign/Makefile ~ sshd/Makefile TAGGED OPENBSD_5_7 > remove roaming support; ok djm@ (markus@) ~ sshconnect2.c ~ sshconnect.c TAGGED OPENBSD_5_7 > fd leaks; report Qualys Security Advisory team; ok deraadt@ (markus@) systat ~ cache.c ~ cache.h ~ pftop.c > fix a number of minor issues in the state view: > 1. peak and rate computations were using uint32 size which caused > the fields from working for states that transferred more than 4GB > also fix an unnecessary cast to double found while looking for the bug. > 2. When creating new cache entries for existing states, start time from > 0 rather than the current time to prevent a "spike" in rate and peak. > (canacar@) tmux ~ input.c ~ tmux.h > A couple of missing printflike attributes, from Andrey Starodubtsev. > (nicm@) ~ environ.c > Clear the environment properly by looping until it is empty rather than > looping over it (which may skip entries), from Brad King. (nicm@) ~ alerts.c ~ cmd-find.c ~ tmux.1 ~ tmux.h > Add hooks for alerts (bell, silence, activity), from Thomas Adam. (nicm@) which ~ which.1 ~ which.c > Use _PATH_DEFPATH if no PATH in the environment when invoked as which. > (millert@) ~ which.c > Avoid potential read of one byte before the start of a malloc()ed > buffer. From Max Fillinger. (millert@) ~ which.c > Check the return value of snprintf() for potential overflow instead > of doing a manual check beforehand. (millert@) == usr.sbin ========================================================== 10/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.sbin cron ~ atrun.c ~ cron.c ~ crontab.c ~ database.c ~ structs.h > When caching the mtime of the spool directory and system crontab files, > stash a struct timespec, not just a time_t. Fixes a bug where cron > could skip re-reading the spool after two consecutive changes. (millert@) dhcrelay ~ packet.c > "Check UDP length for short as well as long values" - apply the fix from > usr.sbin/dhcpd/packet.c r1.4 to dhcrelay. ok deraadt@ (sthen@) eigrpctl ~ eigrpctl.8 ~ eigrpctl.c ~ parser.c ~ parser.h > Introduce the 'eigrpctl clear neighbors' command. (renato@) eigrpd ~ neighbor.c ~ parse.y ~ rde_dual.c > Order the routing instances by AF and then by AS. > This commit only improves the readability of the 'eigrpctl' show commands. > Additionaly, we can simplify the nbr_compare() and rt_compare() functions > because we know that under no circumstances nbrs or routes from different > AFs or ASes will be in the same tree. Each instance has its own trees of > neighbors and routes. (renato@) ~ eigrp.h ~ interface.c ~ packet.c ~ rde.c > Avoid the use of inet_pton() whenever possible. (renato@) ~ eigrp.h ~ eigrpe.h ~ packet.c ~ tlv.c > Split TLV constants into subcomponents to simplify the code. (renato@) ~ eigrpe.c ~ eigrpe.h ~ hello.c ~ update.c > Do not set the EoT flag in the last startup update. > Unfortunately we don't have enough information to support the NSF feature, > the EIGRP draft is missing a lot of information in this regard. > The EoT flag is used as part of EIGRP NSF to announce the end of the > startup process with a neighbor. The problem is that, when we set this > flag, the Cisco neighbors will assume that we support NSF, which is > not true. When this happens, these routers expect us to do things that > we don't know, and weird things can happen. So let's play safe and just > ignore this flag for now. (renato@) ~ control.c ~ eigrpd.h ~ eigrpe.c ~ eigrpe.h ~ hello.c ~ interface.c ~ neighbor.c ~ rde_dual.c ~ rtp.c > Add support for manually clearing neighbors. (renato@) ~ eigrpd.h ~ eigrpe.c ~ kroute.c ~ rde_dual.c ~ util.c > Minor tweaks. (renato@) ~ interface.c > Fix bug that happened when a passive interface was shut down and then > reactivated. (renato@) ~ packet.c > Several fixes in the Conditionally Received (CR) mode. > * If we fail to accept a packet when in the CR mode, do not ack it; > * Add missing ntohl() when storing the next multicast sequence tlv; > * The CR flag in the neighbor should be cleared out after use. (renato@) ~ rde_dual.c > Fix detection of the Stuck-in-Active (SIA) state. > We were rearming the SIA and active timer over and over again. Now, > when the maximum of three SIA-Queries are sent, we stop and wait for the > active timeout. (renato@) ~ rde.c ~ rde_dual.c > Better handle explicit nexthops. > If an explicit nexthop was advertised for a route, show it in the > 'eigrpctl show topology' command output instead of the address of the > advertising neighbor (implicit nexthop). (renato@) ~ rde_dual.c > For each prefix, order routes by their nexthop. (renato@) hostapd ~ print-802_11.c > Remove wireless turbo mode support. It is a non-standard extension > which only worked with ath(4) devices from a decade ago. > Diff tested on ath(4) hardware by me to verify that 11a/b modes still work. > ok reyk deraadt chris sthen kettenis (stsp@) installboot ~ landisk_installboot.c > sync() needs unistd.h (tobiasu@) ldapd ~ control.c ~ ldapd.c > Properly remove unix sockets (control & listening) upon exit of the > parent process. > Child process was killed by pledge because it tried to remove the > control socket and didnt have cpath - anyway it couldnt remove it since > it had chrooted.. > ok jmatthew@ deraadt@ (landry@) lpr ~ common_source/common.c ~ common_source/displayq.c ~ common_source/lp.h ~ common_source/rmjob.c ~ lpd/printjob.c > Move prototypes of local functions from lp.h to the .c files and make > functions static if possible. Move delay() to lpd/printjob.c and fix > an annoying typo. > ok deraadt@ (tb@) ndp ~ ndp.c > Do not print an error if the list of prefixes is empty. > This is becoming more and more comon now that the list is only > used for autoconfigured prefix. > From Florian Riehm. (mpi@) ntpd ~ ntpd.c > sneaky whitespace snuck in again (deraadt@) relayd ~ check_tcp.c ~ hce.c ~ relayd.h > This adds the host_error output and the http code (when available) to the > host-check log. > ok claudio@ (benno@) smtpd - filter_api.3 > remove filter api man page, this needs rework > ok gilles (jung@) ~ enqueue.c ~ parse.y ~ util.c > Let smtpd start on machines without a FQDN as hostname. > Ok millert@ gilles@ jung@ (sunil@) ~ mta.c > Check whether a string from imsg is empty by dereferencing the char* > rather than using strlen. That way we don't have to walk through the > entire string. > ok sunnil@ (mmcc@) sysmerge ~ sysmerge.8 ~ sysmerge.sh > No more /etc/mail/aliases.db. (ajacoutot@) tcpdump ~ print-802_11.c > Remove wireless turbo mode support. It is a non-standard extension > which only worked with ath(4) devices from a decade ago. > Diff tested on ath(4) hardware by me to verify that 11a/b modes still work. > ok reyk deraadt chris sthen kettenis (stsp@) ~ nfsfh.h ~ parsenfsfh.c ~ print-nfs.c > Remove preprocessor conditions (checking for ancient OSs) that try to > divine whether a file handle was generated on the current host. More > simplifications to come. > supported by deraadt@ (mmcc@) ~ parsenfsfh.c > remove a macro that was made an identity function by my previous commit > ok deraadt@ (mmcc@) vmctl ~ vmctl.c > Update the comments in vmctl.c that are describing function arguments. > (reyk@) vmd ~ control.c > remove a duplicate break statement (jsg@) ~ vmm.c > Update the comment that is decribing terminate_vm()'s arguments, > sort the description of arguments for init_emulated_hw(). > Pointed out my mlarkin (reyk@) ~ pci.c ~ virtio.c > stdio.h is not needed here anymore. (mlarkin@) ~ loadfile_elf.c ~ virtio.c ~ vmd.h ~ vmm.c > vmd(8) sometimes attempts page-crossing data copies between the host > and guest. The readpage/writepage ioctls of vmm(4) do not support this > and they return EINVAL on such attempts since recently. > Avoid page-crossing guest memory accesses by changing read_page() and > write_page() into read_mem() and write_mem() that can copy arbitrary > lengths of data between host<->guest without page-crossing accesses. > This also allows us to remove page-wise copy-loops in a few places. > ok mlarkin@ (stefan@) =============================================================================== _______________________________________________ owc mailing list [email protected] http://www.squish.net/mailman/listinfo/owc
