[odc] Daily src changes for 2007-12-02

[ODC]

OpenBSD src changes summary for 2007-12-02
==========================================

share/man                               sys/arch/aviion/aviion
sys/arch/aviion/include                 sys/arch/luna88k/include
sys/arch/luna88k/luna88k                sys/arch/m88k/conf
sys/arch/m88k/include                   sys/arch/m88k/m88k
sys/arch/mvme68k/include                sys/arch/mvme88k/include
sys/arch/mvme88k/mvme88k                sys/dev/acpi
sys/dev/pci                             sys/dev/usb
sys/net                                 usr.sbin/acpidump

== share ============================================================= 01/03 ==

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

man

  ~ man4/pf.4                             

  > DIOC{GET,ADD}STATE incorrectly use a user provided pointer without using
  > copyin/out.  Change the API so that the state is included in the ioctl
  > argument, so the ioctl wrappers take care of copying memory as appropriate.
  > Also change the DIOCGETSTATE API to be more useful.  Instead of getting
  > an arbitrarily "numbered" state (using numbering that can change between
  > calls), instead search based on id and creatorid.  If you want to monitor
  > only a particular state, you can now use the bulk functions first to find
  > the appropriate id/creatorid and then fetch it directly from then on.
  > ok dlg@ henning@ (pascoe@)

== sys =============================================================== 02/03 ==

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

arch/aviion/aviion

  ~ locore.S                              

  > Do not pass UPAGES and USPACE (under the name USIZE) in assym.h, code which
  > needs it includes <machine/param.h> already. (miod@)

arch/aviion/include

  + mplock.h                              

  > Rework the __mp_lock code to not spin at spllock(), kinda similar to the
  > x86 __mp_lock changes, but keeping the internal __cpu_simplelock_t to
  > guarantee atomic access to the __mp_lock fields. (miod@)

arch/luna88k/include

  + mplock.h                              

  > Rework the __mp_lock code to not spin at spllock(), kinda similar to the
  > x86 __mp_lock changes, but keeping the internal __cpu_simplelock_t to
  > guarantee atomic access to the __mp_lock fields. (miod@)

arch/luna88k/luna88k

  ~ locore.S                              

  > Do not pass UPAGES and USPACE (under the name USIZE) in assym.h, code which
  > needs it includes <machine/param.h> already. (miod@)

arch/m88k/conf

  ~ files.m88k                            

  > Rework the __mp_lock code to not spin at spllock(), kinda similar to the
  > x86 __mp_lock changes, but keeping the internal __cpu_simplelock_t to
  > guarantee atomic access to the __mp_lock fields. (miod@)

arch/m88k/include

  ~ asm.h                                 ~ asm_macro.h

  > Sort and clean definitions. No functional changes. (miod@)

  ~ lock.h                                

  > Provide a specific rw_cas() function for MP kernels, simulating a really
  > atomic compare-and-swap operation. (miod@)

  + mplock.h                              

  > Rework the __mp_lock code to not spin at spllock(), kinda similar to the
  > x86 __mp_lock changes, but keeping the internal __cpu_simplelock_t to
  > guarantee atomic access to the __mp_lock fields. (miod@)

arch/m88k/m88k

  ~ pmap.c                                ~ subr.S

  > Provide faster bcopy() and bzero() routines for pmap_copy_page() and
  > pmap_zero_page(). (miod@)

  ~ m88100_machdep.c                      

  > Change DAE processing so that the C code no longer needs the EF_xxx defines
  > from assym.h. (miod@)

  ~ m88k_machdep.c                        

  > Provide a specific rw_cas() function for MP kernels, simulating a really
  > atomic compare-and-swap operation. (miod@)

  ~ trap.c                                

  > Do not take biglock for NOLOCK system calls. (miod@)

  ~ m88110_mmu.S                          

  > Remove 88110 control registers accessors which are never used anywhere.
  > (miod@)

  ~ m88k_machdep.c                        

  > When setting up the vectors page, do not put a nop as the first instruction
  > if the processor is a 88110, since only the 88100 suffers from the ``will
  > fetch one instruction too far'' bug. (miod@)

  ~ eh_common.S                           

  > Since the 88110 doesn't disable the FPU when handling an exception (and
  > does
  > not need to), do not try to handle exceptions occuring when we re-enable
  > shadowing as special - these are just nested exceptions.
  > While there, add a workaround for the 88110 rte errata (#18). (miod@)

  ~ trap.c                                

  > Only check for errata #16 for instruction faults. (miod@)

  ~ eh_common.S                           ~ genassym.cf

  > Do not pass UPAGES and USPACE (under the name USIZE) in assym.h, code which
  > needs it includes <machine/param.h> already. (miod@)

  + mplock.c                              

  > Rework the __mp_lock code to not spin at spllock(), kinda similar to the
  > x86 __mp_lock changes, but keeping the internal __cpu_simplelock_t to
  > guarantee atomic access to the __mp_lock fields. (miod@)

  ~ eh_common.S                           ~ m88110_fp.S
  ~ vectors_88110.S                       

  > The beginning of a real floating-point exception handler for the 88110. The
  > existing code to enable TCFP was broken, as it was not setting the TCFP bit
  > in the right register.
  > So far, the exception handler will deliver SIGFPE in all cases. It will
  > eventually do the necessary rounding, and handle the odd-numbered register
  > pair operation, as I get time to write this (or see how much can be lifted
  > from the 88100 floating-point exception code). (miod@)

arch/mvme68k/include

  ~ intr.h                                

  > One last missing IPL_VM. (miod@)

arch/mvme88k/include

  ~ m88110.h                              

  > Remove 88110 control registers accessors which are never used anywhere.
  > (miod@)

  ~ mplock.h                              

  > Rework the __mp_lock code to not spin at spllock(), kinda similar to the
  > x86 __mp_lock changes, but keeping the internal __cpu_simplelock_t to
  > guarantee atomic access to the __mp_lock fields. (miod@)

arch/mvme88k/mvme88k

  ~ m88410.c                              

  > Better inline assembler constructs. (miod@)

  ~ locore.S                              

  > Do not pass UPAGES and USPACE (under the name USIZE) in assym.h, code which
  > needs it includes <machine/param.h> already. (miod@)

  ~ locore.S                              

  > Add a workaround for the MVME197LE reset issue when not system controller.
  > (miod@)

  ~ m197_machdep.c                        

  > The beginning of a real floating-point exception handler for the 88110. The
  > existing code to enable TCFP was broken, as it was not setting the TCFP bit
  > in the right register.
  > So far, the exception handler will deliver SIGFPE in all cases. It will
  > eventually do the necessary rounding, and handle the odd-numbered register
  > pair operation, as I get time to write this (or see how much can be lifted
  > from the 88100 floating-point exception code). (miod@)

  ~ m88110.c                              

  > Rework mmu initialization so that we do not actually enable translation
  > until cmmu_set_sapr(). Also, do not enable snooping on MVME197LE, so that
  > we don't have to add workarounds for snooping problems later. (miod@)

dev/acpi

  ~ acpi.c                                

  > avoid trailing spaces on output; marco@ ok. (fgsch@)

  ~ acpi.c                                

  > fix the trailing space for real now. (fgsch@)

  ~ dsdt.c                                

  > Added AMLOP_BREAKOPINT function handler for X40 systems
  > ok marco@ (jordan@)

dev/pci

  ~ agp_i810.c                            

  > Put back some of the handling for writes to the GATT on intel
  > agp bridges. This got lost in the merge before.
  > Tested by Peter Hessler on amd64.  ok tedu (oga@)

  ~ pcidevs                               

  > Add Radeon Mobility M10 NT, as seen on his T42p by Alexander Holupirek, who
  > kindly looked it up on www.pcidatabase.com. (krw@)

  ~ pcidevs.h                             ~ pcidevs_data.h

  > Sync (krw@)

dev/usb

  ~ uaudio.c                              

  > spelling fix in debug printf (robert@)

net

  ~ pf_ioctl.c                            

  > DIOCADDSTATE would always dereference a NULL pointer during this copy
  > and what it was copying would get overwritten anyway.  Remove the copy
  > and avoid a panic.
  > DIOCGETSTATE would incorrectly dereference a pointer to a pointer,
  > causing another panic.  Fix this. (pascoe@)

  ~ pf_ioctl.c                            

  > Set expiry timestamp when importing a state, otherwise it expires on the
  > next expiry run.
  > ok dlg@ henning@ (pascoe@)

  ~ pf_ioctl.c                            

  > Don't leak state if key allocation fails during add.
  > ok dlg@ henning@ (pascoe@)

  ~ pf_ioctl.c                            

  > Don't put state key if pf_insert_state fails.  pf_detach_state would have
  > put it for us already.
  > Also, fix cut-n-paste error in previous commit.
  > ok dlg@ henning@ (pascoe@)

  ~ pf_ioctl.c                            

  > initialize altq->altq_disc to NULL, from Max Laier (dhartmei@)

  ~ pf_ioctl.c                            

  > When pf_insert_state state succeeds, increase the state count on the
  > default rule.
  > When pf_insert_state fails, it's because a matching state already exists.
  > Return a better error code to the user in this case.
  > ok henning@ dlg@ (pascoe@)

  ~ pf_ioctl.c                            ~ pfvar.h

  > DIOC{GET,ADD}STATE incorrectly use a user provided pointer without using
  > copyin/out.  Change the API so that the state is included in the ioctl
  > argument, so the ioctl wrappers take care of copying memory as appropriate.
  > Also change the DIOCGETSTATE API to be more useful.  Instead of getting
  > an arbitrarily "numbered" state (using numbering that can change between
  > calls), instead search based on id and creatorid.  If you want to monitor
  > only a particular state, you can now use the bulk functions first to find
  > the appropriate id/creatorid and then fetch it directly from then on.
  > ok dlg@ henning@ (pascoe@)

== usr.sbin ========================================================== 03/03 ==

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

acpidump

  ~ asl_dump.c                            

  > Fixes acpidump on DL145 systems; method defined within another method
  > ok brad@ (jordan@)

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