[odc] Daily src changes for 2007-09-07

ODC auto at squish.net
Sat Sep 8 07:00:01 BST 2007


OpenBSD src changes summary for 2007-09-07
==========================================

lib/libc                                sbin/restore
share/man                               sys/arch/amd64/stand/biosboot
sys/arch/i386/i386                      sys/arch/i386/include
sys/arch/i386/isa                       sys/arch/i386/pci
sys/arch/i386/stand/biosboot            sys/dev/acpi
sys/dev/i2c                             sys/dev/ic
sys/dev/pci                             sys/dev/pcmcia
sys/dev/usb                             sys/kern
sys/net                                 sys/net80211
sys/scsi                                sys/sys
sys/ufs/ffs                             sys/uvm
usr.bin/cvs                             usr.bin/man
usr.bin/nc                              usr.bin/sendbug
usr.bin/sudo                            usr.sbin/eeprom
usr.sbin/hoststatectl                   usr.sbin/hoststated
usr.sbin/repquota                       usr.sbin/rtsold
usr.sbin/syslogd                        

== lib =============================================================== 01/06 ==

  http://www.openbsd.org/cgi-bin/cvsweb/src/lib

libc

  ~ stdio/fgets.3                         

  > Say NUL when referring to \0 (cloder@)

  ~ sys/socket.2                          

  > IPv6 is not an ARPA protocol;
  > from Robert Elz, netbsd-bugs pr#36933
  > ok otto deraadt (jmc@)

== sbin ============================================================== 02/06 ==

  http://www.openbsd.org/cgi-bin/cvsweb/src/sbin

restore

  ~ tape.c                                

  > check fgets return value
  > use sizeof buf instead of TP_BSIZE
  > use strcspn to properly overwrite '\n' in fgets returned buffer
  > ok moritz@ ray@ (chl@)

== share ============================================================= 03/06 ==

  http://www.openbsd.org/cgi-bin/cvsweb/src/share

man

  ~ man9/malloc.9                         

  > Document M_ZERO. (art@)

  ~ man4/iwn.4                            

  > two reasons to put the ultra long URL in a display instead of (the usual)
  > .Pa:
  > (1) it will stop newer versions of groff complaining
  > (2) it avoids line split so you can cut/paste it more easily (jmc@)

  ~ man4/pci.4                            

  > remove a double "Intel Intel";
  > use this commit as an excuse to remove more .Tn sludge; (jmc@)

  ~ man4/nfe.4                            

  > nfe supports MCP73; diff from brad, confirmed by ckuethe (jmc@)

  ~ man4/Makefile                         + man4/adl.4

  > adl(4) (kettenis@)

  ~ man4/iic.4                            

  > Xr adl 4 (deraadt@)

== sys =============================================================== 04/06 ==

  http://www.openbsd.org/cgi-bin/cvsweb/src/sys

arch/amd64/stand/biosboot

  ~ biosboot.S                            

  > Comment fix.  It's 0x7F ELF, not 0xFF ELF. (weingart@)

arch/i386/i386

  ~ powernow-k7.c                         

  > Based upon a diff orignally submitted by devin smith, Do two things run the
  > check that the maxfid and the highest reported acpi p_state frequency agree
  > only
  > after actually getting the acpi p_state. Secondly prefer the legacy
  > powernow
  > table over ACPI because machines of this vintage of are likely to have
  > broken
  > ACPI implementations.
  > Tested by devin smith and tilo stritzky.
  > ok beck@ (gwk@)

  ~ ipifuncs.c                            ~ lapic.c

  > Remove some left-overs from the TSC based microtime. We don't need
  > to synchronize the tsc between CPUs anymore. While here, also remove
  > the slow TLB IPI since it's been dead for a while.
  > noticed by mickey
  > toby@ ok (art@)

  ~ esm.c                                 ~ i686_mem.c
  ~ k6_mem.c                              ~ machdep.c
  ~ mem.c                                 ~ mpbios.c
  ~ via.c                                 

  > Use M_ZERO in a few more places to shave bytes from the kernel.
  > eyeballed and ok dlg@ (art@)

arch/i386/include

  ~ cpu.h                                 ~ intr.h
  ~ intrdefs.h                            

  > Remove some left-overs from the TSC based microtime. We don't need
  > to synchronize the tsc between CPUs anymore. While here, also remove
  > the slow TLB IPI since it's been dead for a while.
  > noticed by mickey
  > toby@ ok (art@)

arch/i386/isa

  ~ isa_machdep.c                         

  > Use M_ZERO in a few more places to shave bytes from the kernel.
  > eyeballed and ok dlg@ (art@)

arch/i386/pci

  ~ ichpcib.c                             

  > Modify the heuristic for detection of speedstep to recognize a compatible
  > mobile Intel Pentium 3 as found in Dell c610 fixes pr5565.
  > ok beck, toby (gwk@)

  ~ glxsb.c                               ~ pci_intr_fixup.c

  > Use M_ZERO in a few more places to shave bytes from the kernel.
  > eyeballed and ok dlg@ (art@)

arch/i386/stand/biosboot

  ~ biosboot.S                            

  > Comment fix.  It's 0x7F ELF, not 0xFF ELF. (weingart@)

dev/acpi

  ~ acpiprt.c                             

  > Fix code that figures out the PCI bus number for a _PRT configuration
  > object.
  > This should get rid of the "multiple bus 0" problem seen on some machines,
  > and fix PCI interrupt routing on them.
  > ok marco@ (a while ago) (kettenis@)

dev/i2c

  ~ files.i2c                             + asc7621.c

  > Enter adl(4), a driver for the Andigilog aSC7621 written by Mike Belopuhov.
  > (kettenis@)

  ~ asc7621.c                             

  > missing scale factor; from Mike Belopuhov (deraadt@)

dev/ic

  ~ ar5xxx.c                              

  > use M_ZERO (reyk@)

  ~ mpi.c                                 

  > take advantage of the new M_ZERO malloc flag. (dlg@)

  ~ siop.c                                ~ trm.c

  > The obvious bzero/memset -> M_ZERO changes. (krw@)

  ~ aic79xx.c                             ~ aic79xx_openbsd.c
  ~ aic7xxx.c                             ~ aic7xxx_openbsd.c

  > The obvious bzero/memset -> M_ZERO changes. (krw@)

  ~ rt2560.c                              ~ rt2661.c

  > use new malloc M_ZERO flag to shrink kernel.
  > remove <malloc.h> from files where malloc is not used. (damien@)

dev/pci

  ~ ahc_pci.c                             

  > The obvious bzero/memset -> M_ZERO changes. (krw@)

  ~ if_ipw.c                              ~ if_iwi.c
  ~ if_wpi.c                              ~ if_iwn.c
  ~ if_nfe.c                              

  > use new malloc M_ZERO flag to shrink kernel.
  > remove <malloc.h> from files where malloc is not used. (damien@)

  ~ if_iwn.c                              ~ if_iwnreg.h

  > make sure the length of the first segment of a Tx descriptor is a multiple
  > of 4 by inserting padding bytes when necessary.
  > 802.11 QoS headers have a length that is not a multiple of 4. (damien@)

  ~ if_iwn.c                              

  > properly acknowledge interrupts.  this fixes the interrupts storm i was
  > seeing with vmstat -i.  re-enable interrupts coalescing. (damien@)

dev/pcmcia

  ~ if_xe.c                               ~ pcmcia.c
  ~ pcmcia_cis.c                          

  > sprinkle some M_ZERO love. (fgsch@)

dev/usb

  ~ if_ral.c                              ~ if_rum.c
  ~ if_zyd.c                              ~ if_uath.c
  ~ ueagle.c                              

  > use new malloc M_ZERO flag to shrink kernel.
  > remove <malloc.h> from files where malloc is not used. (damien@)

kern

  ~ kern_malloc.c                         

  > Add the long requested M_ZERO flag to malloc(9).
  > But the reason for this isn't some kind of "we can make it use the
  > pre-zeroed pages and zero the freelist in the idle loop and OMG I can
  > has optimisatiuns" which would require tons of infrastructure and make
  > everything slower.
  > The reason is that it shrinks other code. And that's good.
  > dlg@ ok, henning@ ok (before he read the diff) (art@)

  ~ kern_kthread.c                        ~ kern_proc.c
  ~ kern_sysctl.c                         ~ subr_autoconf.c
  ~ subr_disk.c                           ~ sys_generic.c
  ~ sysv_sem.c                            ~ sysv_shm.c
  ~ tty.c                                 ~ tty_nmea.c
  ~ tty_pty.c                             ~ tty_subr.c
  ~ uipc_usrreq.c                         ~ vfs_init.c
  ~ vfs_subr.c                            ~ vfs_syscalls.c

  > Use M_ZERO in a few more places to shave bytes from the kernel.
  > eyeballed and ok dlg@ (art@)

net

  ~ if_trunk.c                            

  > use M_ZERO (reyk@)

  ~ rtsock.c                              

  > Check if rtm_hdrlen is smaller then the passed message or it would be
  > possible
  > to access uninitialised memory.
  > Set dst to 0 on error, the error path tries to access dst but dst is inited
  > later down the code. This fixes a kernel panic seen by aanriot@
  > OK henning@ (claudio@)

  ~ pf.c                                  

  > Do not recalculate TCP payload length in pf_test_rule() as it has
  > been cached in pd->p_len.
  > ok henning@ markus@ (bluhm@)

net80211

  ~ ieee80211_node.c                      

  > use new malloc M_ZERO flag + replace MALLOC with malloc (damien@)

scsi

  ~ cd.c                                  ~ ch.c
  ~ safte.c                               ~ scsi_ioctl.c
  ~ ses.c                                 

  > The obvious bzero/memset -> M_ZERO changes. (krw@)

sys

  ~ malloc.h                              

  > Add the long requested M_ZERO flag to malloc(9).
  > But the reason for this isn't some kind of "we can make it use the
  > pre-zeroed pages and zero the freelist in the idle loop and OMG I can
  > has optimisatiuns" which would require tons of infrastructure and make
  > everything slower.
  > The reason is that it shrinks other code. And that's good.
  > dlg@ ok, henning@ ok (before he read the diff) (art@)

ufs/ffs

  ~ ffs_vfsops.c                          

  > Use M_ZERO in a few more places to shave bytes from the kernel.
  > eyeballed and ok dlg@ (art@)

uvm

  ~ uvm_amap.c                            ~ uvm_aobj.c
  ~ uvm_stat.h                            ~ uvm_swap.c

  > Use M_ZERO in a few more places to shave bytes from the kernel.
  > eyeballed and ok dlg@ (art@)

== usr.bin =========================================================== 05/06 ==

  http://www.openbsd.org/cgi-bin/cvsweb/src/usr.bin

cvs

  ~ client.c                              ~ server.c

  > Do not use global option -V (GNU cvs doesn't understand it), be very
  > verbose per default on server-side instead.
  > OK joris@ (tobias@)

  ~ checkout.c                            ~ client.c

  > Added support for single file checkout.
  > OK joris@ (tobias@)

  ~ checkout.c                            ~ rcs.c
  ~ update.c                              

  > Sprinkle some magic so checkout/update -p works
  > in both local and server mode.
  > OK tobias@ (joris@)

  ~ cvs.c                                 ~ file.c

  > Do not fatal if getpwuid() fails, instead ignore user configuration files.
  > OK joris@ (tobias@)

  ~ log.c                                 ~ log.h
  ~ update.c                              

  > Properly send out rcs info header with "E " in server mode (stderr in
  > local mode). Format style change also enables opencvs to be used with
  > cvsweb (+ adjustment).
  > OK joris@ (tobias@)

man

  ~ man.1                                 

  > we have man section 3f by default now; (jmc@)

nc

  ~ atomicio.c                            ~ atomicio.h

  > Synced atomicio implementation in nc and sendbug with ssh.
  > OK djm@, joris@, ray@ (tobias@)

sendbug

  ~ atomicio.c                            

  > Synced atomicio implementation in nc and sendbug with ssh.
  > OK djm@, joris@, ray@ (tobias@)

sudo

  ~ configure                             ~ configure.in

  > fix a configure bug introduced in the last commit (millert@)

== usr.sbin ========================================================== 06/06 ==

  http://www.openbsd.org/cgi-bin/cvsweb/src/usr.sbin

eeprom

  ~ eeprom.8                              ~ main.c

  > some small cleanup, including standard synopsis/usage and sort options;
  > ok fgsch deraadt (jmc@)

hoststatectl

  ~ hoststatectl.c                        

  > simple style change in the switch (reyk@)

  ~ hoststatectl.8                        ~ hoststatectl.c
  ~ parser.c                              ~ parser.h

  > add a "show sessions" command to dump the complete list of running relay
  > sessions. (reyk@)

hoststated

  ~ hoststated.h                          ~ log.c

  > add a function to print delays in hours, minutes, and seconds (reyk@)

  ~ parse.y                               

  > store the table's port as the relay's dstport (reyk@)

  ~ hoststated.h                          

  > remove unused flags field from the structure (reyk@)

  ~ control.c                             ~ hoststated.h
  ~ pfe.c                                 ~ relay.c

  > add an interface to dump running relay sessions to the control socket
  > (reyk@)

repquota

  ~ repquota.c                            

  > Fix use of arguments in usage().
  > ok winiger (mbalmer@)

rtsold

  ~ rtsol.c                               

  > unexpeced -> unexpected;
  > from Wouter Schoot, netbsd-bugs #36931 (jmc@)

syslogd

  ~ syslogd.c                             

  > Initialize length variable slen properly to avoid possible EINVAL
  > errors from setsockopt(2) in function double_rbuf().
  > ok markus@ mpf@ (bluhm@)

===============================================================================


More information about the odc mailing list