[odc] Daily src changes for 2004-06-09
ODC
auto at squish.net
Thu Jun 10 08:00:43 BST 2004
OpenBSD src changes summary for 2004-06-09
==========================================
distrib/sets gnu/egcs
gnu/usr.bin/cvs include/limits.h
sbin/isakmpd sbin/ping6
sbin/route sys/arch/alpha/include
sys/arch/amd64/amd64 sys/arch/hppa/hppa
sys/arch/hppa/include sys/arch/i386/eisa
sys/arch/i386/i386 sys/arch/i386/isa
sys/arch/i386/pci sys/arch/m88k/include
sys/arch/powerpc/include sys/arch/vax/include
sys/kern sys/sys
sys/uvm usr.sbin/bgpd
usr.sbin/rdate usr.sbin/tcpdump
== distrib =========================================================== 01/06 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/distrib
sets
~ lists/man/mi
> sync (deraadt@)
~ lists/base/md.luna88k ~ lists/base/md.mvme88k
~ lists/comp/md.luna88k ~ lists/comp/md.mvme88k
~ lists/man/md.luna88k ~ lists/man/md.mvme88k
> binutils sync (miod@)
== gnu =============================================================== 02/06 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/gnu
egcs
~ gcc/config/m88k/m88k.c
> More meaningful formulation of the __va_stk initialization in
> __builtin_saveregs(); no functional change. (miod@)
~ gcc/config/m88k/m88k.c
> In FUNCTION_ARG_ADVANCE(), do not pad the register position counter if the
> parameter is going to hit the stack. (miod@)
~ gcc/config/m88k/m88k.c
> The GO_IF_LEGITIMATE_ADDRESS fix in m88k.h, as well as all the FUNCTION_ARG
> and FUNCTION_ARG_ADVANCE fixes in m88k.c, allow the optimized bcopy
> sequences to be reliable again, so enable them back. (miod@)
usr.bin/cvs
~ src/commit.c ~ src/cvs.h
~ src/filesubr.c ~ src/history.c
~ src/server.c ~ src/wrapper.c
+ lib/xsize.h
> Several potential security problems found and fixed by Stefan Esser &
> Sebastian Krahmer.
> ok millert@ (otto@)
~ src/commit.c ~ src/cvs.h
~ src/filesubr.c ~ src/history.c
~ src/server.c ~ src/wrapper.c
+ lib/xsize.h
TAGGED OPENBSD_3_5
> MFC:
> Fix by otto@
> Several potential security problems found and fixed by Stefan Esser &
> Sebastian Krahmer. (brad@)
~ src/commit.c ~ src/cvs.h
~ src/filesubr.c ~ src/history.c
~ src/server.c ~ src/wrapper.c
+ lib/xsize.h
TAGGED OPENBSD_3_4
> MFC:
> Fix by otto@
> Several potential security problems found and fixed by Stefan Esser &
> Sebastian Krahmer. (brad@)
== include =========================================================== 03/06 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/include
limits.h
~ limits.h
> According to POSIX, _POSIX_PATH_MAX should be 256 but _POSIX_SYMLINK_MAX
> should only be 255. You would think they would be the same but you'd be
> wrong. (millert@)
== sbin ============================================================== 04/06 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sbin
isakmpd
~ message.c
> only accept DELETEs during an authenticated INFORMATIONAL exchange.
> Fix for recent problem disclosed by Thomas Walpuski.
> ok ho@ (hshoexer@)
~ conf.c ~ exchange.c
~ ike_phase_1.c ~ ike_quick_mode.c
~ ipsec.c ~ isakmp_cfg.c
~ message.c ~ pf_key_v2.c
~ transport.c ~ udp.c
> Style nits. hshoexer@ ok (ho@)
~ message.c
TAGGED OPENBSD_3_5
> MFC:
> Fix by hshoexer@
> only accept DELETEs during an authenticated INFORMATIONAL exchange.
> Fix for recent problem disclosed by Thomas Walpuski. (brad@)
~ message.c
TAGGED OPENBSD_3_4
> MFC:
> Fix by hshoexer@
> only accept DELETEs during an authenticated INFORMATIONAL exchange.
> Fix for recent problem disclosed by Thomas Walpuski. (brad@)
ping6
~ ping6.c
> correct mistake in usage(). Chris Pinnock (itojun@)
route
~ route.c
> Fix IPv4 name->address translation. Addresses like "10.1000" will not be
> accepted anymore, but constructs like "route add 10.1.2/24 <gw>" will
> finally do the right thing. ok millert@ (cedric@)
== sys =============================================================== 05/06 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sys
arch/alpha/include
~ pmap.h
> rename POOLPAGE macros to pmap_map_direct
> break out uvm_km_page bits for this case, no thread here
> lots of testing tech@, deraadt@, naddy@, mickey@, ... (tedu@)
arch/amd64/amd64
~ cpu.c
> Minor sync from the SMP branch.
> NVRAM_RESET_SOFT -> NVRAM_RESET_JUMP
> ok deraadt@ (art@)
arch/hppa/hppa
~ pmap.c
> rename POOLPAGE macros to pmap_map_direct
> break out uvm_km_page bits for this case, no thread here
> lots of testing tech@, deraadt@, naddy@, mickey@, ... (tedu@)
arch/hppa/include
~ pmap.h
> rename POOLPAGE macros to pmap_map_direct
> break out uvm_km_page bits for this case, no thread here
> lots of testing tech@, deraadt@, naddy@, mickey@, ... (tedu@)
arch/i386/eisa
~ eisa_machdep.c
> pull in KNF from SMP (deraadt@)
arch/i386/i386
~ apic.c ~ mpbios.c
TAGGED SMP
> we do not use bitmask_snprintf; ok grange art (deraadt@)
~ machdep.c
TAGGED SMP
> we trust the calibrated pentium_mhz more, so don't overwrite it unless
> we know the cpu speed has changed. fixes pr3814 from mark pecaut. (tedu@)
~ ioapic.c
TAGGED SMP
> Do not call lapic_set_softvectors, or lapic_set_lvt, if no ioapics are
> present
> in the system. ok art@ (drahn@)
~ locore.s
TAGGED SMP
> Do the DDB prologue for kcopy too. Fix copy_fault to deal with the changed
> stack frame layout. Fixes the error case of copy{in,out}
> We run too high quality software, we never tickle the error case.
> ok markus@ deraadt@ (art@)
arch/i386/isa
~ nvram.h
> Minor sync from the SMP branch.
> NVRAM_RESET_SOFT -> NVRAM_RESET_JUMP
> ok deraadt@ (art@)
arch/i386/pci
~ elan520.c
> fix compilation for !I586_CPU case; ok grange@ (djm@)
arch/m88k/include
~ va-m88k.h
> Among other tricky (or nice, depending upon your involvment) features,
> the 88Open ABI allows arguments to be passed through registers, then on
> the stack, the through registers, then on the stack again, with some
> function signatures.
> Unfortunately, gcc has not been designed with this possibility in mind,
> and does not handle this situation optimally (since no sane architecture
> does this anyway) and will reserve stack storage for the second set of
> in-registers parameters.
> Fixing this is quite non-trivial (trust me on this), so in the meantime
> keep at wasting space stack, but teach varargs about this so that it can
> compensate correctly. (miod@)
arch/powerpc/include
~ pmap.h
> rename POOLPAGE macros to pmap_map_direct
> break out uvm_km_page bits for this case, no thread here
> lots of testing tech@, deraadt@, naddy@, mickey@, ... (tedu@)
arch/vax/include
~ pmap.h
> rename POOLPAGE macros to pmap_map_direct
> break out uvm_km_page bits for this case, no thread here
> lots of testing tech@, deraadt@, naddy@, mickey@, ... (tedu@)
kern
~ init_main.c ~ kern_clock.c
~ kern_fork.c ~ kern_ktrace.c
~ kern_subr.c ~ kern_synch.c
~ kern_sysctl.c ~ kern_time.c
> Merge in a piece of the SMP branch into HEAD.
> Introduce the cpu_info structure, p_cpu field in struct proc and global
> scheduling context and various changed code to deal with this. At the
> moment no architecture uses this stuff yet, but it will allow us slow and
> controlled migration to the new APIs.
> All new code is ifdef:ed out.
> ok deraadt@ niklas@ (art@)
~ vfs_default.c
> in theory, vnlock should alays be NULL in the generic lock (sic) functions.
> in reality, sometimes it's not. we don't trust vnlock, and since it's 100%
> guaranteed to panic if it gets here, just completely stop using it.
> crash by henning, ok deraadt (tedu@)
sys
~ proc.h
TAGGED SMP
> spacing (deraadt@)
~ kernel.h ~ proc.h
~ sched.h
TAGGED SMP
> Merge in a piece of the SMP branch into HEAD.
> Introduce the cpu_info structure, p_cpu field in struct proc and global
> scheduling context and various changed code to deal with this. At the
> moment no architecture uses this stuff yet, but it will allow us slow and
> controlled migration to the new APIs.
> All new code is ifdef:ed out.
> ok deraadt@ niklas@ (art@)
uvm
~ uvm_km.c
> rename POOLPAGE macros to pmap_map_direct
> break out uvm_km_page bits for this case, no thread here
> lots of testing tech@, deraadt@, naddy@, mickey@, ... (tedu@)
== usr.sbin ========================================================== 06/06 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.sbin
bgpd
~ control.c ~ session.c
~ session.h
> move to a dynamically allocated struct pollfd array.
> we used a ststic one with OPEN_MAX entries, which is a rather arbitary
> limit
> as OPEN_MAX is _not_ the max # of open fds we can have, but just a default
> for that setting.
> in the same move we have to allocate the peer_l array, basically there
> for pfd-index to peer pointers to prevent peer list scans all time,
> dynamiccaly to. we overallocate a little and use that reserve until we
> have to realloc again later to prevent reallocs for every single control
> connection or a single flapping peer.
> help & ok claudio (henning@)
rdate
~ ntp.c
> * Bring rdate's SNTP support into compliance with SNTPv4 (RFC 2030).
> * More robust handling of NTP error conditions (e.g. host or
> service unreachable).
> * Improve the detection of stale and/or spoofed NTP responses
> from servers.
> * Add support for getaddrinfo(3)'s multiple host support if
> error conditions occur (e.g. round-robin DNS, and the first
> NTP server isn't responding, try the next host in line).
> * Minor formatting/code cleanup.
> ok henning@ (alexander@)
tcpdump
~ privsep.c
> Take better care of people down under living in X and a half hour
> timezones.
> ok canacar@ (otto@)
===============================================================================
More information about the odc
mailing list