wpaul [Sat, 28 Oct 2000 09:00:20 +0000 (09:00 +0000)]
Add PCI IDs for some additional cardbus cards. Yes, there really is
a RealTek 8139 cardbus device. Unfortunately it doesn't quite work yet
because the CIS parser barfs on it.
Submitted by msmith, with some small tweaks by me.
bde [Sat, 28 Oct 2000 08:20:33 +0000 (08:20 +0000)]
Quick fix for namespace problems which broke many ports. Spell
uint16_t as "unsigned short" like it used to be and uint32_t as
"unsigned int" not like it used to be (was: "unsigned long").
msmith [Sat, 28 Oct 2000 07:32:13 +0000 (07:32 +0000)]
Some AML code assumes that a function without an explicit ReturnOp will
return the last value returned by a nested method call. This violates
the ACPI spec, but is implemented by the Microsoft interpreter, and thus
vendors can (and do) get away with it.
Intel's stance is that this is illegal and should not be supported.
As they put it, however, we have to live in the real world. So go ahead
and implement it.
msmith [Sat, 28 Oct 2000 07:29:08 +0000 (07:29 +0000)]
FreeBSD-specific customisations:
- Set debugger options for kernel build
- Define some missing functions
- Bring in GCC defines
- Disable the 'wbinvd' macro as it conflicts with our inline
msmith [Sat, 28 Oct 2000 07:26:48 +0000 (07:26 +0000)]
Add some extra fuctions:
- AcpiGetProcessorID (fetch the ACPI processor ID for a given ACPI_HANDLE)
- AcpiSetSystemSleepState (set the Sx sleeping state, proposed by Intel
but not actually implemented)
msmith [Sat, 28 Oct 2000 07:07:02 +0000 (07:07 +0000)]
Add some verbiage about the new ACPICA code. Note that this probably
breaks building this config, as the old ACPI code will conflict. The
old code will be going away shortly, so this should not be an issue.
msmith [Sat, 28 Oct 2000 06:59:48 +0000 (06:59 +0000)]
Initial FreeBSD OSPM (operating system power management) modules for
ACPICA. Most of these are still works in progress. Support exists for:
- Fixed feature and control method power, lid and sleep buttons.
- Detection of ISA PnP devices using ACPI namespace.
- Detection of PCI root busses using ACPI namespace.
- CPU throttling and sleep states (incomplete)
- Thermal monitoring and cooling control (incomplete)
- Interface to platform embedded controllers (mostly complete)
- ACPI timer (incomplete)
- Simple userland control of sleep states.
- Shutdown and poweroff.
msmith [Sat, 28 Oct 2000 05:01:06 +0000 (05:01 +0000)]
Initial import of the Intel-maintained ACPI Component Architecture. This
is Intel's reference implementation of the core operating-system ACPI
support. (This import is on the vendor branch.)
jhb [Sat, 28 Oct 2000 00:28:15 +0000 (00:28 +0000)]
The x86 atomic operations are already locked, so they do not need an
additional locked instruction to guarantee a write barrier for the acquire
variants.
wollman [Fri, 27 Oct 2000 20:50:14 +0000 (20:50 +0000)]
Initialize rn_mklist in rn_newpair(). The undocumented assumption
seems to be that the nodes are bzero'd beforehand, but the submitter
found that this was not always the case, and in any event defensive
programming here costs epsilon squared.
PR: 22244
Submitted by: Dave Gillam <daveg@chiaro.com>
dfr [Fri, 27 Oct 2000 09:36:34 +0000 (09:36 +0000)]
Backout revision 1.7 which was a bad idea since it would force people
to reinstall boot1 after a 'make world'.
Unfortunately this means that people who have already installed a new
boot1 from a 'make world' after 2000/09/18 *must* reinstall it after
their next build using something like:
ru [Fri, 27 Oct 2000 07:19:17 +0000 (07:19 +0000)]
Fetch the protocol header (TCP, UDP, ICMP) only from the first fragment
of IP datagram. This fixes the problem when firewall denied fragmented
packets whose last fragment was less than minimum protocol header size.
Found by: Harti Brandt <brandt@fokus.gmd.de>
PR: kern/22309
jhb [Fri, 27 Oct 2000 02:59:30 +0000 (02:59 +0000)]
- Use MUTEX_DECLARE() and MTX_COLD for the WITNESS code's internal mutex so
it can function before malloc(9) is up and running.
- Add two new options WITNESS_DDB and WITNESS_SKIPSPIN. If WITNESS_SKIPSPIN
is enabled, then spin mutexes are ignored by the WITNESS code. If
WITNESS_DDB is turned on and DDB is compiled into the kernel, then the
kernel will drop into DDB when either a lock hierarchy violation occurs
or mutexes are held when going to sleep.
- Add some new sysctls:
debug.witness_ddb is a read-write sysctl that corresponds to WITNESS_DDB.
The kernel option merely changes the default value to on at boot.
debug.witness_skipspin is a read-only sysctl that one can use to determine
if the kernel was compiled with WITNESS_SKIPSPIN.
- Wipe out the BSD/OS-specific lock order lists. We get to build our own
lists now as we add mutexes to the kernel.
green [Fri, 27 Oct 2000 01:19:02 +0000 (01:19 +0000)]
Fix problems people were having with large -O levels with GCC and
getting libutil/libcrypt to work properly. I've determined that GCC
thinks it can inline all functions, including weak-symboled ones, if
it feels like it.
Create a new stub.c and move any stubs there to prevent inlining.
Thanks to jdp and William S. Duncanson for helping me finally find the
problem.
wpaul [Fri, 27 Oct 2000 00:15:04 +0000 (00:15 +0000)]
Yet another bug fix/optimization for the Davicom DM9100/9102: increase
the PCI latency timer value to 0x80. Davicom's Linux driver does this,
and it drastically reduces the number of TX underruns in my tests. (Note:
this is done only for the Davicom chips. I'm not sure it's a good idea to
do it for all of them.)
Again, still waiting on confirmation before merging to stable.
ru [Thu, 26 Oct 2000 13:14:48 +0000 (13:14 +0000)]
RFC 791 says that IP_RF bit should always be zero, but nothing
in the code enforces this. So, do not check for and attempt a
false reassembly if only IP_RF is set.
Also, removed the dead code, since we no longer use dtom() on
return from ip_reass().