[odc] Daily src changes for 2005-10-09
ODC
auto at squish.net
Mon Oct 10 08:00:37 BST 2005
OpenBSD src changes summary for 2005-10-09
==========================================
lib/libc libexec/ld.so
share/man sys/arch/amd64/amd64
sys/arch/i386/conf sys/arch/macppc/dev
sys/arch/macppc/include sys/arch/macppc/macppc
sys/arch/macppc/stand sys/arch/powerpc/include
sys/arch/powerpc/powerpc sys/dev/pci
sys/net sys/net80211
usr.bin/cvs usr.bin/rcs
== lib =============================================================== 01/05 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/lib
libc
~ sys/sigaction.2
> ANSI function prototypes in examples; ok otto (jaredy@)
== libexec =========================================================== 02/05 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/libexec
ld.so
~ dlfcn.c ~ library.c
~ library_mquery.c ~ library_subr.c
~ loader.c ~ resolve.c
~ resolve.h
> introduce object ref count macros (suggested by dale). no functional
> change. (kurt@)
~ library_subr.c
> fix incomplete ref count check in _dl_link_dlopen (use macro now). (kurt@)
== share ============================================================= 03/05 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/share
man
~ man4/lge.4
> don't bother mentioning VLAN support and tweak checksumming part. (brad@)
~ man5/hosts.equiv.5
> log in(v.); (jmc@)
== sys =============================================================== 04/05 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sys
arch/amd64/amd64
~ vector.S
> spelling (fgsch@)
arch/i386/conf
~ GENERIC
> what the hell, enable ipmi (deraadt@)
arch/macppc/dev
~ openpic.c
> While running pending interrupts, enable external interrupt, this way
> low priority interrupts do not starve high priority. (drahn@)
~ if_bmreg.h
> remove unused NETWORK_BUFSIZE. (brad@)
~ if_bm.c
> remove some NetBSD/OpenBSD portability cruft.
> compiles ok kettenis@ (brad@)
arch/macppc/include
~ powerpc.h
> Nearly functional crashdump support for macppc. Because savecore
> does not recognize the resulting crashdumps, the writing has been disabled.
> Better here than in my forest of trees. (drahn@)
~ cpu.h
> Provide a machdep sysctl to determine if altivec is avaliable on macppc
> (drahn@)
arch/macppc/macppc
~ autoconf.c ~ clock.c
~ machdep.c
> Nearly functional crashdump support for macppc. Because savecore
> does not recognize the resulting crashdumps, the writing has been disabled.
> Better here than in my forest of trees. (drahn@)
~ machdep.c
> Subtle changes to the powerpc fpu handling, basically fpuproc is handled
> in a non-raceable manner inside save_fpu and enable_fpu so that the
> eventual SMP case will not grab a half loaded fpu context. (drahn@)
~ cpu.c ~ machdep.c
> Provide a machdep sysctl to determine if altivec is avaliable on macppc
> (drahn@)
arch/macppc/stand
~ boot.c
> Try to recover part of the reserved space from the OF_claim above. (drahn@)
arch/powerpc/include
~ pmap.h
> Nearly functional crashdump support for macppc. Because savecore
> does not recognize the resulting crashdumps, the writing has been disabled.
> Better here than in my forest of trees. (drahn@)
~ fpu.h
> Subtle changes to the powerpc fpu handling, basically fpuproc is handled
> in a non-raceable manner inside save_fpu and enable_fpu so that the
> eventual SMP case will not grab a half loaded fpu context. (drahn@)
arch/powerpc/powerpc
~ pmap.c
> Nearly functional crashdump support for macppc. Because savecore
> does not recognize the resulting crashdumps, the writing has been disabled.
> Better here than in my forest of trees. (drahn@)
~ fpu.c ~ process_machdep.c
~ trap.c ~ vm_machdep.c
> Subtle changes to the powerpc fpu handling, basically fpuproc is handled
> in a non-raceable manner inside save_fpu and enable_fpu so that the
> eventual SMP case will not grab a half loaded fpu context. (drahn@)
~ trap.c
> unbreak altivec support. (drahn@)
dev/pci
~ if_lge.c
> use pci_matchbyid(). (brad@)
~ if_lge.c ~ if_nge.c
> fix bus_dmamem_alloc() failure case. (brad@)
~ if_ti.c
> some code cleaning..
> - function return value on separate line
> - ANSI protos
> - remove return at end of void function
> - splimp -> splnet (brad@)
~ if_em.c
> remove colon after "address" when printing the MAC address. (brad@)
~ if_tireg.h
> remove ETHER_CRC_LEN and ETHER_HDR_LEN (brad@)
~ if_bgereg.h ~ if_tireg.h
> special case sparc64 for now and lower the Jumbo slots
> down from 384 to 54. (brad@)
~ if_bge.c
> - move bus_dmamap_load_mbuf() to where it should be in bge_newbuf_std()
> - add comments mentioning where we're re-using mbufs (brad@)
~ if_bge.c ~ if_bgereg.h
> BGE_RSLOTS is only used in one spot, BGE_TX_RING_CNT
> should be used instead. (brad@)
net
~ if_trunk.c
> use label (reyk@)
~ if_bridge.c ~ if_bridge.h
> if the wireless hostap interface belongs to a bridge, learn the node's
> address dynamically on this interface after successful association.
> this could speed wireless roaming to openbsd accesspoints.
> ok jason@, jsg@ can't see anything obectionable (reyk@)
net80211
~ ieee80211_node.c
> if the wireless hostap interface belongs to a bridge, learn the node's
> address dynamically on this interface after successful association.
> this could speed wireless roaming to openbsd accesspoints.
> ok jason@, jsg@ can't see anything obectionable (reyk@)
== usr.bin =========================================================== 05/05 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.bin
cvs
~ rcs.c
> fix 2 off-by-one's which was causing us a whole load of crap;
> ok niallo@ (joris@)
~ log.c
> fix cvs_log() so it doesn't segfault when called before
> cvs_command is set.
> problem found by & discussed with Nuno Morgadinho thanks! (joris@)
rcs
~ ci.c
> - sync checkin_usage()
> - add support for checking-in to arbitrary revisions via -r option.
> eg:
> ci -r1.30 <file>
> check in file, jump to revision 1.30 and set that revision to HEAD.
> (niallo@)
~ ci.1
> - document `-r' switch as a single option instead of two distinct options.
> - sync with checkin_usage()
> input from jmc@ (niallo@)
~ ci.c
> simplify if() conditions;
> ok niallo@ (joris@)
~ ci.c
> only get the log message when it's not specified on the command-line;
> noticed by niallo@ (joris@)
~ rcsdiff.c
> if there is a problem generating the diff for one of the files, continue
> to process other files but eventually exit with status 2. this mimics
> GNU rcs behaviour.
> ok joris@ (niallo@)
~ ci.c
> if -r is not specified, we need to set newrev to file->rf_head.
> simpler fix than what i had in mind pointed out by joris. (niallo@)
===============================================================================
More information about the odc
mailing list