Commit Graph

1288 Commits

Author SHA1 Message Date
ed
322413c46c Add memmove() to the kernel, making the kernel compile with Clang.
When copying big structures, LLVM generates calls to memmove(), because
it may not be able to figure out whether structures overlap. This caused
linker errors to occur. memmove() is now implemented using bcopy().
Ideally it would be the other way around, but that can be solved in the
future. On ARM we don't do add anything, because it already has
memmove().

Discussed on:	arch@
Reviewed by:	rdivacky
2009-02-28 16:21:25 +00:00
raj
50b3c45b0e Prefer register usage style to be more consistent with the rest of the
trap_subr.S code.
2009-02-27 12:18:17 +00:00
raj
b86fc67839 Make Book-E debug register state part of the PCB context.
Previously, DBCR0 flags were set "globally", but this leads to problems
because Book-E fine grained debug settings work only in conjuction with the
debug master enable bit in MSR: in scenarios when the DBCR0 was set with
intention to debug one process, but another one with MSR[DE] set got
scheduled, the latter would immediately cause debug exceptions to occur upon
execution of its own code instructions (and not the one intended for
debugging).

To avoid such problems and properly handle debugging context, DBCR0 state
should be managed individually per process.

Submitted by:	Grzegorz Bernacki gjb ! semihalf dot com
Reviewed by:	marcel
2009-02-27 12:08:24 +00:00
nwhitehorn
28bdf866e3 Fix comment: we write the trap vector to SPRG3, not SPRG0. 2009-02-23 19:31:48 +00:00
thompsa
6b0018e885 Change over the usb kernel options to the new stack (retaining existing
naming). The old usb stack can be compiled in my prefixing the name with 'o'.
2009-02-23 18:34:56 +00:00
nwhitehorn
78b3ec3583 Add Altivec support for supported CPUs. This is derived from the FPU support
code, and also reducing the size of trapcode to fit inside a 32 byte handler
slot.

Reviewed by:	grehan
MFC after:	2 weeks
2009-02-20 17:48:40 +00:00
raj
fe8e0abcf5 Additional features for the tsec(4) Ethernet driver.
- interrupt coalescing
  - polling
  - jumbo frames
  - multicast
  - VLAN tagging

The enhanced version of the chip (eTSEC) can also take advantage of:

  - TCP/IP checksum calculation h/w offloading

Obtained from:	Freescale, Semihalf
2009-02-17 14:57:05 +00:00
thompsa
c24b826e84 Add uslcom to the build too.
Reminded by:	Michael Butler
2009-02-15 23:40:29 +00:00
thompsa
15cccb8286 Switch over GENERIC kernels to USB2 by default.
Tested by:	make universe
2009-02-15 22:33:44 +00:00
nwhitehorn
2cc6f5c8ec Add support for the I2S and davbus audio controllers found in Apple PowerPC
hardware.

Submitted by:	Marco Trillo
2009-01-25 18:20:15 +00:00
nwhitehorn
d8a0e03d76 Fix a race condition where interrupts set up after boot could be enabled in
the PIC before the interrupt handler was set. If the interrupt triggered in
that window, then the interrupt vector would be disabled.

Reported by:	Marco Trillo
2009-01-25 17:50:53 +00:00
nwhitehorn
ba6a8e1391 Fix a race condition in kiic(4) made possible by the way the device's STOP
condition is sent. We used to put the bus in the STOP state, but returned
without waiting for that to actually occur.

Submitted by:	Marco Trillo
2009-01-20 14:09:12 +00:00
nwhitehorn
70abc2d871 Provide a device description for macio-attached ATA cells. 2009-01-19 23:25:18 +00:00
nwhitehorn
d9b4dc27b5 Driver for Apple Keywest I2C controllers found in MacIO ASICs. Used for
power and thermal control, as well as GPIOs on Xserves and controlling
sound codecs for Apple built-in audio.

Submitted by:	Marco Trillo
Obtained from:	NetBSD
2009-01-15 02:52:31 +00:00
raj
a481347383 Clean up BookE low-level exceptions code.
Improve comments, fix style(9) and typos, unify separators.

Obtained from:	Freescale, Semihalf
2009-01-13 16:19:58 +00:00
raj
923b601516 Clean up BookE pmap.
Improve comments, eliminate redundant debug output, fix style(9) and other
minor tweaks for code readability.

Obtained from:	Freescale, Semihalf
2009-01-13 16:15:49 +00:00
raj
b729364e00 Rework BookE pmap towards multi-core support.
o Eliminate tlb0[] (a s/w copy of TLB0)
  - The table contents cannot be maintained reliably in multiple MMU
    environments, where asynchronous events (invalidations from other cores)
    can change our local TLB0 contents underneath.
  - Simplify and optimize TLB flushing: system wide invalidations are
    performed using tlbivax instruction (propagates to other cores), for
    local MMU invalidations a new optimized routine (assembly) is introduced.

o Improve and simplify TID allocation and management.
  - Let each core keep track of its TID allocations.
  - Simplify TID recycling, eliminate dead code.
  - Drop the now unused powerpc/booke/support.S file.

o Improve page tables management logic.

o Simplify TLB1 manipulation routines.

o Other improvements and polishing.

Obtained from:	Freescale, Semihalf
2009-01-13 15:41:58 +00:00
nwhitehorn
87790de246 Some early Macintosh GPIO controllers don't provide reg properties for
interrupt-only GPIOs. Honor this, and allow interrupt attachment, but not
read/write access for such devices.

Reported by:	Niels Eliasen
2009-01-12 01:36:01 +00:00
nwhitehorn
c41f2ca87e Add a new quirk type so that the MacIO driver will assign memory resources
belonging to a devices children, in analogy to the way we handle interrupts
for SCC serial devices. This is required to counteract overly deep nesting
on onboard audio devices.

Submitted by:	Marco Trillo
2009-01-06 01:54:57 +00:00
nwhitehorn
4ca12119c3 Fix the OFW interrupt map parser to use its own idea of the number of interrupt
cells in the map, instead of using a value passed to it and then panicing if it
disagrees. This fixes interrupt map parsing for PCI bridges on some Apple
Uninorth PCI controllers.

Reported by:	marcel
Tested on:	G4 iBook, Sun Ultra 5
2009-01-03 19:38:47 +00:00
nwhitehorn
3fcef6d9c2 Modularize the Open Firmware client interface to allow run-time switching
of OFW access semantics, in order to allow future support for real-mode
OF access and flattened device frees. OF client interface modules are
implemented using KOBJ, in a similar way to the PPC PMAP modules.

Because we need Open Firmware to be available before mutexes can be used on
sparc64, changes are also included to allow KOBJ to be used very early in
the boot process by only using the mutex once we know it has been initialized.

Reviewed by:    marius, grehan
2008-12-20 00:33:10 +00:00
raj
95725c6fde Minor spelling fix in E500 locore. 2008-12-18 18:28:03 +00:00
raj
188fbdbc6c Extend and improve MPC85XX Local Bus management.
- Make LBC resources management self-contained: introduce explicit LBC
  resources definition (much like the OCP), provide dedicated rman for LB mem
  space.

- Full configuration of an LB chip select device: program LAW and BR/OR, map
  into KVA, handle all LB attributes (bus width, machine select, ecc,
  write protect etc).

- Factor out LAW manipulation routines into shared code, adjust OCP area
  accordingly.

- Other LBC fixes and clean-ups.

Obtained from:	Semihalf
2008-12-18 18:27:12 +00:00
raj
f811858422 Fix E500 cache invalidation routines.
When invalidating the i/d-cache we need to wait until the core complex is
really finished with the operation.

Obtained from:	Semihalf
2008-12-17 15:54:25 +00:00
raj
898a13418a Rework E500 locore.
- split bootstrap code into more modular routines, which will also be used for
  the non-booting cores
- clean up registers usage
- improve comments to better reflect reality
- eliminate dead or redundant code
- other minor fixes

This refactoring is a preliminary step before importing dual-core (MPC8572)
support.

Obtained from:	Freescale, Semihalf
2008-12-17 15:44:34 +00:00
raj
8bd8077930 Minor clean up of BookE/MPC85XX: iprove naming and style(9). 2008-12-17 15:31:15 +00:00
raj
d71801145e Improve MPC85XX helper routines.
- Move CCSR accessors to the shared MPC85XX area
- Simplify SVR version subfield handling
- Adjust OCP
2008-12-17 15:27:49 +00:00
imp
39a3668dcc AT_DEBUG and AT_BRK were OBE like 10 years ago, so retire them.
Reviewed by:	peter
2008-12-17 06:56:58 +00:00
nwhitehorn
f0e837432d Adapt parts of the sparc64 Open Firmware bus enumeration code (in particular,
the code for parsing interrupt maps) to PowerPC and reflect their new MI
status by moving them to the shared dev/ofw directory.

This commit also modifies the OFW PCI enumeration procedure on PowerPC to
allow the bus to find non-firmware-enumerated devices that Apple likes to add,
and adds some useful Open Firmware properties (compat and name) to the pnpinfo
string of children on OFW SBus, EBus, PCI, and MacIO links. Because of the
change to PCI enumeration on PowerPC, X has started working again on PPC
machines with Grackle hostbridges.

Reviewed by:	marius
Obtained from:	sparc64
2008-12-15 15:31:10 +00:00
nwhitehorn
1d585eec25 Allow OFW syscons to restore itself when the X server exits or there is a VT switch
by redoing the Open Firmware card initialization calls in ofwfb_set_mode(). This
uses the same trick (setting V_ADP_MODECHANGE) to arrange this as machfb(4) and
creatorfb(4).
2008-12-13 20:53:57 +00:00
nwhitehorn
f830e5c602 Add support for a console mouse pointer on Open Firmware syscons.
MFC after:	7.1-RELEASE
2008-12-13 20:13:22 +00:00
nwhitehorn
03bced1c70 Use a static free packet queue instead of using malloc() to allocate new ADB packets.
This fixes some locking problems.
2008-12-13 18:49:01 +00:00
nwhitehorn
0f0a18ff38 Add the ability to control the sleep LED with led(4). Adding this fairly
useless feature gives us a reasonably complete PMU implementation.
2008-12-09 01:01:02 +00:00
nwhitehorn
f35bca43d4 Clean up the mac GPIO interface a little. Also remove bogus copyright
and 3rd license clause.

Submitted by:	Marco Trillo
2008-12-08 03:00:45 +00:00
nwhitehorn
7a058f6255 Accidentally left ADB out of the PowerPC NOTES file during initial import. 2008-12-08 02:38:13 +00:00
nwhitehorn
19d49bd49b Add facilities to pmu(4) to interrogate battery status on Apple PowerPC
laptops. This includes battery presence detection, charging status, current
and voltage readouts, and charge level indication. The sysctl interface
is somewhat ACPI-like.
2008-12-08 02:37:08 +00:00
nwhitehorn
eff63b34b1 Add support for automated reboot after power failure on Apple Core99 machines
(G3 laptops, all G4 machines, early G5s, G5 Xserves). The relevant sysctl
is named dev.pmu.0.server_mode for mental compatibility with Linux.
2008-12-07 00:42:15 +00:00
nwhitehorn
e6bc656165 Fix some nasty race conditions in the VIA-CUDA driver that ended up preventing
my right mouse button and keyboard LEDs from working due to mangled
configuration packets. Fixed several other races and associated problems in the
main ADB stack that were exposed while fixing this.
2008-12-06 23:26:02 +00:00
ed
9286c815e8 Remove "[KEEP THIS!]" from COMPAT_43TTY. It's not really that important.
Sgtty is a programming interface that has been replaced by termios over
the years. In June we already removed <sgtty.h>, which exposes the
ioctl()'s that are implemented by this interface. The importance of this
flag is overrated right now.
2008-12-02 19:09:08 +00:00
marcel
080c5cc425 Unbreak previous commit. 2008-11-22 22:15:34 +00:00
kib
8fad2283b3 Add sv_flags field to struct sysentvec with intention to provide description
of the ABI of the currently executing image. Change some places to test
the flags instead of explicit comparing with address of known sysentvec
structures to determine ABI features.

Discussed with:	dchagin, imp, jhb, peter
2008-11-22 12:36:15 +00:00
kmacy
9d3bb599b1 - bump __FreeBSD version to reflect added buf_ring, memory barriers,
and ifnet functions

- add memory barriers to <machine/atomic.h>
- update drivers to only conditionally define their own

- add lockless producer / consumer ring buffer
- remove ring buffer implementation from cxgb and update its callers

- add if_transmit(struct ifnet *ifp, struct mbuf *m) to ifnet to
  allow drivers to efficiently manage multiple hardware queues
  (i.e. not serialize all packets through one ifq)
- expose if_qflush to allow drivers to flush any driver managed queues

This work was supported by Bitgravity Inc. and Chelsio Inc.
2008-11-22 05:55:56 +00:00
marcel
39fb5ed1d3 Define LDBL_EPSILON, LDBL_MAX and LDBL_MIN as long double constants.
Submitted by:	Andreas Tobler <andreast-list@fgznet.ch>
Reviewed by:	das@
2008-11-16 19:20:29 +00:00
sobomax
ae5a12cdf1 Fix compilation in the case when kernel doesn't have KDB ebabled.
subr_kdb.c still references breakpoint() in this case.
2008-10-30 21:02:00 +00:00
nwhitehorn
5c55d9c1df Fix some possible infinite loops in the ADB code, and remove some hacks
that were inserted in desperation during bring-up. In addition, move ADB bus
enumeration and child attachment to when interrupts are available.
2008-10-30 15:27:13 +00:00
marcel
64f5c79806 Add support for little-endian compilations to this file. 2008-10-30 03:31:33 +00:00
nwhitehorn
a6fcdd6909 DBDMA can transfer a maximum of 64K - 1 bytes per descriptor, as the byte
count field is 16 bits. Inform ATA of this fact.

Reported by:	Marco Trillo
2008-10-28 22:09:30 +00:00
nwhitehorn
3dde3e513b Clean up some magic numbers in the DBDMA code by replacing them with
appropriately defined constants.

Suggested by:	gnn
2008-10-27 23:11:14 +00:00
marcel
322737440f Add support for kernel profiling for both AIM and BookE.
Obtained from:	Juniper Networks, Inc (BookE support).
2008-10-27 02:36:03 +00:00
marcel
3472a37bb7 Remove unused declarations (interrupt_vector_{base|top}). 2008-10-27 01:51:30 +00:00