34c6df3404
Initialize the two new members of "struct consdev" to NULL so that
...
this file compiles again.
2003-03-08 09:47:13 +00:00
0f55132c91
Initialize the two new members of "struct console" to NULL so that
...
this file compiles again.
2003-03-08 09:42:56 +00:00
67cca2386c
Protect uses of MIPS_R5000 with #ifndef ENABLE_MIPS_R3NKK in new code just
...
as the old code does.
2003-03-08 05:18:25 +00:00
33fcc94c6f
Add support for R5k secondary caches, from code Chris Sekiya sent me a long
...
time ago, with small tweaks by me. Since the R5k doesn't do VCE, the pmap
still needs to be whacked for R5kSC CPUs to work correctly, but this is a
start.
2003-03-08 04:43:24 +00:00
12a8e27857
Remove .pdr section which is created by binutils-2.13.2.1
...
on objcopy call to make ECOFF kernels.
2003-03-08 02:58:55 +00:00
585b922ff4
Initialize the new function pointers to NULL, so this compiles again.
2003-03-08 01:18:42 +00:00
59a910ed6a
Move the MTRR option to std.x86_64.
2003-03-08 00:08:46 +00:00
6107a9b605
crank memory disk from 1360 blocks to 1376 blocks
2003-03-08 00:04:36 +00:00
dfb9f2d504
Remove some constant (use value in bus_space_tags).
2003-03-07 18:24:00 +00:00
b486d4a0ff
Add NULL cn_halt and cn_flush entries to consdev.
2003-03-07 13:30:35 +00:00
02de215ad2
Add cpu* at mainbus.
2003-03-07 11:02:50 +00:00
e7e88128f7
Fix compile error with struct consdev change.
2003-03-06 23:47:35 +00:00
d2c1c747ec
Remove last traces of pccons.
2003-03-06 23:21:36 +00:00
167eeeab9c
Update for kernel include changes.
2003-03-06 22:32:07 +00:00
06114d8df6
Don't descend into ../i386/include anymore.
2003-03-06 22:22:55 +00:00
00289f6f52
Put the correct value in the TPR before entering DDB.
2003-03-06 22:17:44 +00:00
51c335ad9a
Fixup DDB entry and register access.
2003-03-06 20:58:09 +00:00
feeb962699
Correct pci1 bus_space tag.
2003-03-06 20:15:24 +00:00
ea9a7f73e0
* No need for an interface attribute on "dma".
...
* Take the "dma" attribute off fhpib and nhpib, so that leaving out "dma"
causes a compile-time error, rather than a run-time crash.
2003-03-06 18:28:52 +00:00
bba7685f5f
Update for consdev changes.
2003-03-06 18:24:52 +00:00
bcc03a47aa
Add NULL cn_halt and cn_flush entries to consdevs. Hi Matt!
2003-03-06 13:09:28 +00:00
9ead47682a
catch up to consdev update
2003-03-06 07:39:34 +00:00
210f48e582
Remove unneeded conditional code.
2003-03-06 07:15:46 +00:00
3c0a5d1f13
Remove #ifdef __HAVE_GENERIC_SOFT_INTERRUPTS switch and old interrupt
...
support. Ixp12x0 port always uses generic soft interrupt.
2003-03-06 06:17:43 +00:00
edcaa77675
insert macro to protect itself against multiple includion.
2003-03-06 06:14:16 +00:00
3dc865b386
Relocate Discovery first thing in initppc (if not at expected address).
...
This implies that if PMON tells us where the GT is, we write that address
in gtbase.
2003-03-06 06:04:21 +00:00
a190651169
Make most of PPC (OEA) bus_space's identical (or as identical as possible).
...
(except mvmeppc).
2003-03-06 05:25:18 +00:00
27e2af69da
Call db_machine_init to set up the ddb ipi vector.
2003-03-06 00:47:00 +00:00
0b8a5fd80f
Adapt to powerpc/bus.h changes.
2003-03-06 00:20:39 +00:00
4e5e707719
* Rework interrupt code
...
* Redirect some include files to x86
* Enable ioapic usage
* Do MP.
2003-03-05 23:55:57 +00:00
6689073e50
When computing the LOADADDR() of .data and .sdata, use:
...
LOADADDR(previous_section) + (ADDR(current_section) - ADDR(previous_section))
to avoid using SIZEOF(), which does not properly take into account
the size of orphans that might be sorted with the section.
2003-03-05 23:54:22 +00:00
ff2281b498
Add preliminary support Marvell (Galileo) Discovery System Controllers.
...
This code was contributed by Allegro Networks.
2003-03-05 22:08:18 +00:00
ea477f4ba7
Forgot dummy printf.
2003-03-05 16:52:16 +00:00
1f3cd4f95f
Be more verbose about unsupported keys.
2003-03-05 16:18:24 +00:00
2a540cd45f
Forward declare struct lwp as well as struct trapframe.
2003-03-05 11:28:14 +00:00
e41c87001a
fix some comments
2003-03-05 10:27:59 +00:00
ba44f83888
Fix from Frederick Bruckman to get current kernels booting again.
2003-03-05 05:56:26 +00:00
92d0498c3d
Remove port-specific procfs_machdep.c (since they are all identical and
...
a better one resides in powerpc/powerpc/).
2003-03-05 05:43:42 +00:00
6d251b3be3
Make AltiVec registers available via ptrace/procfs. Simplify AltiVec
...
processing. Add a "common" procfs_machdep.c for PowerPC platforms.
Even though it is supposed to be port specific, most (if not all)
PowerPC ports can just use the common one.
2003-03-05 05:27:24 +00:00
91ef1f21bf
Make the apic address unsigned, as it should be.
2003-03-04 23:27:32 +00:00
c89161c65f
ioapic address is not actually a pointer, initialize it as uint32_t
2003-03-04 23:21:25 +00:00
e575c4bc69
Fix some fields that did not have explicit types yet.
2003-03-04 23:20:29 +00:00
48aa516ffe
Use read_psl and write_psl.
2003-03-04 22:43:36 +00:00
e1a37a9604
Define read_psl and write_psl as aliases for read_eflags and write_eflags.
2003-03-04 22:43:09 +00:00
b2a3a55d38
Enable the define for EISA MPBIOS support.
2003-03-04 22:19:32 +00:00
d9bd125d50
Make EISA support conditional (on by default on i386).
2003-03-04 22:19:04 +00:00
182b7b7cda
lower acpi interrupt level to IPL_VM.
...
it's the highest level it can use because it uses malloc(9).
XXX probably, it's still higher than it needs.
ok'ed by thorpej.
2003-03-04 13:44:08 +00:00
e0395ed0f1
G/c old macppc grf entries.
2003-03-04 13:29:12 +00:00
889bcccc72
Attach on recent TiBooks.
...
XXX The mixer doesn't work yet.
2003-03-04 13:28:20 +00:00
fb2cebb577
Pass the address of the intrframe to the ext_intr routine.
2003-03-04 08:34:12 +00:00
e1d9426036
Make to define ${BOARDTYPE}
2003-03-04 07:52:21 +00:00
107803a3a3
Re-arrange things in evbppc & powerpc to support OEA-based eval boards
...
in evbppc. OEA-based board(s) to be added later.
2003-03-04 07:50:57 +00:00
9875f9218d
Add some missing volatiles.
2003-03-04 07:48:09 +00:00
23d2066add
Cast the argument passed to vtophys() to a vaddr_t.
2003-03-04 01:10:50 +00:00
a345044000
s/i386_isa_chipset/x86_isa_chipset/
2003-03-04 01:07:36 +00:00
f28dd1acef
s/i386_isa_chipset/x86_isa_chipset/, it's defined on multiple ports, but
...
used in shared code.
2003-03-04 01:06:38 +00:00
f26c7b7e7f
Note a kernel pmap locking problem in pmap_kenter_pa4m().
2003-03-03 22:43:58 +00:00
7b925fa8cf
Use unsigned long long to print msr values.
2003-03-03 22:23:22 +00:00
2934facefc
use CVAROFF.
2003-03-03 22:17:15 +00:00
60ff7764b1
Use pmap_cpu_has_pg_n()
2003-03-03 22:16:54 +00:00
22c3c5973b
The IDT is an array of struct gate_descriptor.
2003-03-03 22:16:20 +00:00
5d9be5f0db
The IDT only contains gate descriptors, so define it as such.
2003-03-03 22:14:15 +00:00
1b64bc0ae7
define pmap_cpu_has_pg_n and pmap_cpu_has_invlpg to make code sharing
...
easier.
2003-03-03 22:13:16 +00:00
3ebdf8bcc3
Make CVAR asm reference macro CVAROFF, taking an offset.
2003-03-03 22:12:22 +00:00
cf1495e079
Install cpuvar.h
2003-03-03 20:26:18 +00:00
6c0c8e1049
Change the pmap to use 512-byte pages as user page table pages instead of
...
PAGE_SIZE pages (4k). An average small program uses ~6 4k pages (24k),
while the same program only uses ~20 512 byte pages (10k), so it will be
a small memory usage improvement. The large improvement will be the upcoming
ability to share page table pages between processes for shared libraries.
Remaining: should be able to give back ptp pages to the system.
2003-03-02 22:19:06 +00:00
767bfa5fb6
microtime.S depends on assym.h
2003-03-02 22:19:05 +00:00
2b65555cf7
Change the pmap to use 512-byte pages as user page table pages instead of
...
PAGE_SIZE pages (4k). An average small program uses ~6 4k pages (24k),
while the same program only uses ~20 512 byte pages (10k), so it will be
a small memory usage improvement. The large improvement will be the upcoming
ability to share page table pages between processes for shared libraries.
Remaining: should be able to give back ptp pages to the system.
2003-03-02 22:19:04 +00:00
31e2cbf0b5
add some ISO C 1995 I18N functions and types:
...
btowc, wctrans, towctrans, wcscoll, wcsxfrm, wctype_t and wctrans_t.
2003-03-02 22:18:11 +00:00
77accc351e
Use the new pmap_kprotect() function instead of pmap_changeprot().
...
The latter is now a pmap internal function; and it wasn't used correctly
in mkclock.c anyway.
2003-03-02 21:42:48 +00:00
22cf03a09f
pmap_changeprot() is not a UVM <-> PMAP interface function; make it internal.
...
Provide pmap_kprotect() for MD code to change protection on a range of
kernel addresses (cuurent users: intr.c and mkclock.c).
2003-03-02 21:37:20 +00:00
7f72a2bff2
Clean up some unneeded "mca.h" and "eisa.h" includes, make one that is
...
needed dependent on !__x86_64__. To be revisited later.
2003-03-02 18:27:14 +00:00
636c7ca859
x86_64 has no mca.h and eisa.h (should perhaps just generate empty ones)
2003-03-02 18:11:03 +00:00
8d69f823b1
Redirect to x86/
2003-03-02 18:09:42 +00:00
8aa243df44
Bump SYMTAB_SPACE to 192000.
2003-03-02 04:42:40 +00:00
9c5dcb82ac
Cleanup G2_LOCK/G2_UNLOCK macros per discussion on port-dreamcast:
...
- Use _cpu_intr_suspend()/_cpu_intr_resume() in G2_LOCK/G2_UNLOCK macros
rather than _cpu_exception_suspend()/_cpu_exception_resume().
TLB miss exceptions could happen on accessing buffer, but it will never
block G2bus cycle anyway.
- Restore the previous SR.IMASK value in G2_UNLOCK().
- some misc cosmetics.
Should fix spontaneous reboots by "ping -s 8192 host" on the LAN Adapter.
2003-03-02 04:23:16 +00:00
5401f85320
Restore MQ to trapframe from mcontext since it's in both.
2003-03-02 01:07:55 +00:00
0cb6724bc7
lock_machdep.c moved here from arch/i386/i386.
2003-03-01 22:55:02 +00:00
f555536d9b
Move lock_machdep.c to arch/x86/x86.
2003-03-01 22:53:39 +00:00
986f7ca902
Add machdep sysctl support. Support booted_device, consdev, and
...
printfataltraps.
2003-03-01 21:51:59 +00:00
aeaf29a3a3
Redirect to x86/cpuvar.h
2003-03-01 18:30:03 +00:00
9ad868eb65
Moved here from i386/include
2003-03-01 18:29:28 +00:00
a3b7772adf
Move the CPU identification out of machdep.c into a seperate file.
2003-03-01 16:37:51 +00:00
de9139c67d
Follow the spec more closely, and do an ljmp immediately after enabling
...
protected mode, not several instructions afterwards.
2003-03-01 16:36:37 +00:00
dbef38adbf
This was still falsely referencing the ipending variable, which isn't
...
used anymore. Use curcpu()->ci_ipending (== cpu_info_primary.ci_pending,
since this is only used in the UP case).
2003-03-01 16:35:07 +00:00
d2315fd2b7
Remove accidentally enabled debug printf.
...
From Enami.
2003-03-01 13:05:37 +00:00
1b5340d327
Add support for large kernels by searching for a physical memory segment
...
to fit it in. As a bonus, the kernel is now also mapped to the virtual
address (i.e. KERNBASE) it is linked at.
2003-03-01 13:01:55 +00:00
5afa6838bf
Do the syscall_plain/syscall_fancy dance on ARM. Shaves a fair number
...
of cycles off the syscall overhead.
Since all COMPAT_LINUX platforms now support __HAVE_SYSCALL_INTERN,
garbage-collect the LINUX_SYSCALL_FUNCTION stuff.
2003-03-01 04:36:38 +00:00
a391bf2e06
Make this play with gcc 3.3. Add -ffreestanding. Cleanup bootxx.
2003-02-28 05:24:56 +00:00
2125ffc454
Move the -D__daddr_t from libsa/Makefile to Makefile.booters so
...
the bootloaders get to benefit from it, not just libsa.
2003-02-27 20:38:34 +00:00
eab869e1c0
Make BRIDGE_IPF an option, and document it. Add it (commented) to GENERIC.
...
Let brconfig tell whether the bridge is using the ipfilter hook, or not.
2003-02-27 19:22:36 +00:00
e6ef3bc041
Some interrupt controllers can select polarity of interrupt signals.
...
add IST_LEVEL_{LOW,HIGH} and IST_EDGE_{FALLING,RISING,BOTH} for them.
2003-02-27 14:55:41 +00:00
33bdc2efe0
Free page table pages to the same VM map they were allocated from.
2003-02-27 14:19:41 +00:00
c03626b07b
Change "i386_trunc_page" to "x86_trunc_page".
2003-02-27 12:22:52 +00:00
f8b0474185
Reinstate some const qualifiers I accidentally removed when moving this
...
file.
2003-02-27 11:22:46 +00:00
229495b6d3
Need to forward declare struct buf and struct pte for function prototypes.
2003-02-27 07:14:19 +00:00
d67ca5b2e7
locator should use physical addresses
2003-02-27 04:06:07 +00:00
dd22310ae1
Add consinit.c
2003-02-27 01:50:40 +00:00
ba3f3ceb24
Catch up with consinit.c move.
2003-02-27 01:49:29 +00:00
b5a7ec69dd
Moved here from arch/i386/i386.
2003-02-27 01:49:05 +00:00
10648a4e04
Moved to x86/x86
2003-02-27 01:48:44 +00:00
d89d6b98f7
Catch up with isa_machdep.c and pci_machdep.c move.
2003-02-27 00:39:40 +00:00
e397aa2f84
Moved here from i386/pci.
2003-02-27 00:30:31 +00:00
b7b9aa9909
Moved to x86/pci.
2003-02-27 00:30:21 +00:00
8556a55a8d
Moved here from i386/isa
2003-02-27 00:28:08 +00:00
2c7dd6813f
Moved to x86/isa.
2003-02-27 00:27:54 +00:00
c0ed70b13c
Adapt for move to x86/include.
2003-02-27 00:12:21 +00:00
20e439b5cb
Move a few more files to x86/include. Trim the list of files to install
...
in /usr/include a bit.
2003-02-27 00:10:56 +00:00
2b14e212b1
Install header files.
2003-02-26 22:27:00 +00:00
ab4edb55ec
Adapt for i386/x86 change.
2003-02-26 22:21:19 +00:00
21d5b938d7
Enable USE_TOPDOWN_VM. This also makes it possible to use large address
...
spaces, so bump MAXDSIZ to 1GB.
2003-02-26 21:54:35 +00:00
f720949c2c
Adapt for move of files to arch/x86.
2003-02-26 21:33:36 +00:00
34cdc350a6
The mementry struct is packed by default now, remove lo/hi usage.
2003-02-26 21:32:20 +00:00
c0f1fc2c3e
Add various include files for MP; mostly redirects to x86/foo.h
2003-02-26 21:31:12 +00:00
1fe3d0e71c
Redirect a few includes to x86/foo.h
2003-02-26 21:29:34 +00:00
a33bcafdf1
Redirect some includes to x86/foo.h
2003-02-26 21:28:59 +00:00
cffd488563
Adapt for move of files to arch/x86.
2003-02-26 21:28:20 +00:00
8375b2d985
Move some files out of i386 into x86, so that they can be shared with
...
other ports.
2003-02-26 21:26:09 +00:00
b9f3bdb8fb
Added L3CR_CONFIG to the parameter list for 745x L3 cache configuration.
2003-02-26 21:14:32 +00:00
0c5117e1f9
Added configuration entries for L3CR_CONFIG for L3 caches on 745x
...
accelerators. Thanks to Monroe Williams.
2003-02-26 21:10:51 +00:00
7109206620
Added L3CR_CONFIG for support of 745x G4 L3 cache configuration.
2003-02-26 21:05:23 +00:00
4eb3db09d9
Move all `ID prom' stuff into the prom library.
2003-02-26 17:39:05 +00:00
fb425a12ca
Re-define `struct memarr' to have the same layout as the V2 ROM memory
...
property to avoid gratuitous copying in prom_makememarr().
Have prom_makememarr() accept a NULL pointer, in which case it will return
the size of the array needed to store the memory descriptors.
2003-02-26 14:25:20 +00:00
f5831d3d87
Store the physical memory descriptors on top of the kernel instead of
...
allocating a fixed sized array.
2003-02-26 14:18:24 +00:00
04ef5573bf
Make common macros like every other port. Now newfs_lfs builds again.
2003-02-26 06:21:09 +00:00
869bbf806d
Add some RAS support. Don't print out a message when we encounter
...
trap instructions.
2003-02-25 23:32:03 +00:00
23ea635a14
This file has a _KERNEL dependency on <machine/cpu.h>. So include it.
...
[GENERIC.MP builds again.]
2003-02-25 23:29:53 +00:00
3ea6b8be93
Add missing splbio() protecting of bufpool access. Add a clarifying
...
comment to <sys/buf.h> reminding everyone of the need for splbio().
2003-02-25 21:25:40 +00:00
eb14e86676
Add a new BUF_INIT() macro which initializes b_dep and b_interlock, and
...
use it. This fixes a few places where either b_dep or b_interlock were
not properly initialized.
2003-02-25 20:35:31 +00:00
9474297c7f
Clear 'big' bit of segment registers before returning to real mode.
2003-02-25 11:17:36 +00:00
b69aef5f3e
* Use fdloadfile() to avoid opening the the underlying device twice.
...
* Cleanup old stuff.
2003-02-25 08:09:30 +00:00
dd87308a05
Note which network boot protocol we're trying.
2003-02-25 08:06:29 +00:00
e3a92065e8
Fix another off-by-one, this time in the current KVA limited as maintained
...
by pmap_growkernel.
2003-02-25 00:48:00 +00:00
4552b31f45
Clear 'big' bit of descriptors (%ss in particular) before switching to
...
16bit mode. Otherwise the 32bit %esp and (probably) %ebp are used for
stack accesses. (stupid cpu design!)
(approved by christos)
2003-02-24 23:23:14 +00:00
a8bd073e6f
pv_unlink4m(): don't forget to link in a new PV entry after alias checking.
2003-02-24 15:14:51 +00:00
4f493fc42d
Zero bss in start, not in initppc.
2003-02-24 08:06:42 +00:00
8609ed56b8
Only define KERN_AS= object if IDENT has -DLKM.
2003-02-24 07:15:40 +00:00
83642bd3a6
Cast VM_DEFAULT_ADDRESS() to a caddr_t.
2003-02-24 00:49:43 +00:00
20c4b7b844
Change pcb32_pagedir to a paddr_t (after all, it's used as a paddr_t
...
everywhere in the code).
2003-02-23 23:40:01 +00:00
a5781fdcdf
Add LFS support for i386. Tested with a "newfs_lfs -A -O 256" root
...
filesystem.
2003-02-23 23:25:50 +00:00
3a93f29238
Update for LFSv2 support. Ports that have "bootxx_lfs"-style
...
bootblocks currently only support LFSv2, but the second level
"boot" program supports both LFS versions.
2003-02-23 23:23:07 +00:00
681f702ca1
In pmap_physpage_alloc() and pmap_physpage_free(): only the DIAGNOSTIC
...
check is inside the simple_lock()/simple_unlock() pair, so move the locks
inside the #ifdef as well.
Additionally, change the #ifdef from DIAGNOSTIC to DEBUG; the condition
is too weird to be worth checking at DIAGNOSTIC, according ot Jason.
2003-02-23 22:41:24 +00:00
eca5e4743f
Bump ramdisk image size from 1444k to 1500k.
2003-02-23 22:30:46 +00:00
860ed88533
#include opt_uvm.h in machine/vmparam.h (for those ports offering the
...
topdown option) so that including it directly before including
uvm/uvm_param.h (or uvm/uvm_extern.h which includes uvm/uvm_param.h)
and attempting to use topdown won't result in a compiler error.
Problem noted in private email.
2003-02-23 19:13:43 +00:00
b699c38934
Cast the result of VM_DEFAULT_ADDRESS to caddr_t, to make this compile again.
2003-02-23 19:08:18 +00:00
b09955adaf
Merged mac68k and macppc keayboards, since they are the same. The file
...
could move to <arch/mac> if one day we create it.
Fixed french keymap: option was not mapped correctly, thus making
impossible to use |{}[] (tested).
Added jp keymap from PR/15438, and sf, sv, de and uk keymaps from OpenBSD.
(all untested).
2003-02-23 18:09:25 +00:00
013ba73637
Fix a tyop in a comment.
2003-02-23 03:37:40 +00:00
11e36a204f
For x86_64, split the bootinfo memory entry words to avoid growth of the
...
structure due to alignment.
2003-02-23 02:45:44 +00:00
0ee308ef7e
Fix up pmap_growkernel a bit to avoid off-by-one errors.
2003-02-23 02:44:44 +00:00
49a9f4f666
Use BIOS memory info properly. Don't try to allocate areas > 4G out of
...
the IO extent, since it only spans 32 bits.
2003-02-23 02:43:25 +00:00
d399d9df43
fix incorrect interrupt mask handling.
...
pci_imask[IPL_NET] is incorrectly ORed on imask[IPL_SOFTSERIAL].
imask[IPL_NET] should be ORed.
2003-02-22 11:13:10 +00:00
4b074775af
Added a french keyboard layout.
2003-02-22 08:33:56 +00:00
4e8142fde2
correct physical/virtual address handling
...
- to identify device instance, using hardware address.
- when console accesses device, using statically mapped address.
- when tty accesses device, using handler given by bus_space_map().
2003-02-22 05:32:00 +00:00
007c1ca9cf
Remove the assumption in the sun4/sun4c bootstrap code that the kernel
...
is always mapped using PMEGs and SMEGs starting at #0 .
2003-02-21 19:07:36 +00:00
b7f550c62d
Allow the kernel to be loaded at its link address (aka KERNBASE) by the
...
boot program.
2003-02-21 19:04:07 +00:00
b03aef6efc
witch from xor r,r,r to li r,0 to set a register to 0. It's clearer (and
...
better for G4 processors).
2003-02-21 15:14:08 +00:00
008754f2e6
Increase symbol table space - it did not fit anymore.
2003-02-21 08:55:37 +00:00
c1234c6385
Add TOPDOWN VM support.
2003-02-21 03:41:52 +00:00
a7a7697279
Don't use dv_unit to determine console. Back to comparing iobase again.
...
We always assume ixpcom is at statically mapped address (0xf0000000).
2003-02-21 01:53:35 +00:00
6248ef9b7e
implement ioctl
2003-02-21 00:31:08 +00:00
203288830a
Convert a few types into things that are more accurate, mostly:
...
int's to unsigned int/u_int where they shouldn't go negative.
int's to boolean_t's where they're being used as bools.
No real functional change (in the produced asm a few condition codes changed)
2003-02-21 00:23:03 +00:00
df0a9badc6
Introduce "top down" memory management for mmap()ed allocations. This
...
means that the dynamic linker gets mapped in at the top of available
user virtual memory (typically just below the stack), shared libraries
get mapped downwards from that point, and calls to mmap() that don't
specify a preferred address will get mapped in below those.
This means that the heap and the mmap()ed allocations will grow
towards each other, allowing one or the other to grow larger than
before. Previously, the heap was limited to MAXDSIZ by the placement
of the dynamic linker (and the process's rlimits) and the space
available to mmap was hobbled by this reservation.
This is currently only enabled via an *option* for the i386 platform
(though other platforms are expected to follow). Add "options
USE_TOPDOWN_VM" to your kernel config file, rerun config, and rebuild
your kernel to take advantage of this.
Note that the pmap_prefer() interface has not yet been modified to
play nicely with this, so those platforms require a bit more work
(most notably the sparc) before they can use this new memory
arrangement.
This change also introduces a VM_DEFAULT_ADDRESS() macro that picks
the appropriate default address based on the size of the allocation or
the size of the process's text segment accordingly. Several drivers
and the SYSV SHM address assignment were changed to use this instead
of each one picking their own "default".
2003-02-20 22:16:05 +00:00
b8081a591a
Change locking prerequisites for the pv_changepte4m() helper function
...
to avoid locking recursion.
2003-02-20 16:22:49 +00:00
3cd76da95a
Add vlan(4) pseudo device. Approved by martin@
2003-02-20 13:30:44 +00:00
c3ed92df3e
Add a few new devices.
2003-02-20 12:02:31 +00:00
b42cde14b7
Set KERN_AS=obj now that LKMs work. (otherwise when you loaded LKMs they
...
wouldn't have all the library routines available).
2003-02-19 23:55:48 +00:00
b66bce23df
Finish PV table locking. Terminology and locking strategy stolen from
...
the alpha port.
2003-02-19 22:27:08 +00:00
b7f4f30f48
Revert previous (in line with <i386/mtrr.h> reversion).
2003-02-19 05:38:58 +00:00
d3fce65758
Reserve the I/O ports used by the CRT controller to keep them from
...
being allocated for other purposes such as PCMCIA I/O space. Fixes
port-i386/5177.
2003-02-19 04:07:56 +00:00
feaeb437d4
Make COMPAT_NETBSD32 compile and work again (though not with SAs).
2003-02-19 00:37:33 +00:00
807c14b1f2
Undo constification of mtrr_funcs. I'll deal with it later.
2003-02-19 00:18:34 +00:00
f2ef2f16fa
Revert previous. This file is also used on x86_64, you see.
2003-02-19 00:15:19 +00:00
d286e74769
pgt_page_alloc: use uvm_km_kmemalloc() so we can honor the PR_NOWAIT flag
...
pmap_kenter_pa4m: lock the pmap we're working on.
2003-02-18 22:05:08 +00:00
d2309c1fda
add file-system SMBFS and pseudo-device nsmb, uncommented, marked as
...
experimental
2003-02-18 20:13:03 +00:00
2b390d41d0
Move makememarr() into promlib.c.
2003-02-18 13:36:51 +00:00
51762449f8
add nsmb device-major
2003-02-18 12:26:50 +00:00
60a45bbcd7
add missing networking options, commented.
...
ok'd by ross.
2003-02-18 05:50:51 +00:00
9a38c06746
sync with GENERIC
...
add: options SYSTRACE
2003-02-17 22:44:37 +00:00
4b8928ad4a
files.ixp12x0
...
no need device ixpcom in evbarm/conf/files.evbarm move it to
arm/ixp12x0/files.ixp12x0
ixp12x0_com.c:
some fix around address handling
1. Do not call bus_space_map() in ixpcominit(). Calling bus_space_map()
is not safe here, because bus_space_map() calls uvm_km_valloc() but
uvm is not yet initialized.
2. Use dv_unit to determine console instead comparering iobase.
Now you can attach ixpcom0 with physical address like this:
ixpcom* at ixpsip? addr 0x90000000 size 0x4000
Statically mapped address (0xf0000000) is still usable.
ixp12x0_clk:
1. access PLL_CFG register via bus_space
2. Make the delay() working correctly. (bug fix)
3. Start the timer device without interrupt on attach time.
Now delay() called before cpu_initclocks() works fine.
ixp12x0_pci:
1.Mapping PCI type0/1 configuration space to the upper address.
2."PCI I/O Cycle Access" mapping to same virtual address(VA==PA)
but size of this mapping increase to 1MByte because fails
cause couldnt set L2 table.
3.use bus_space address handling in ixp12x0_pci.c.
2003-02-17 20:51:52 +00:00
7a3efea0bf
Make whitespace in multi-way loops line up a bit nicer.
2003-02-17 12:32:13 +00:00
8883b06039
No need to protect headers with #ifdef _KERNEL if they're never installed
...
in /usr/include.
2003-02-17 11:35:01 +00:00
e9d72b5b52
Make ALLOC_TRACE work again.
2003-02-17 08:21:39 +00:00
8f123087a2
Undo accidental commit of this file.
2003-02-16 17:42:13 +00:00
519f8f3d30
Add uax(4) (and url(4) in some cases).
2003-02-16 17:35:13 +00:00
c502753d44
Fix pasto in previous.
2003-02-15 13:30:11 +00:00
84e3fcabb3
Redo part of rev. 1.237; the pages at KERNBASE might not be mapped when the
...
kernel has just been loaded.
2003-02-15 13:17:41 +00:00
0ba0225830
Initialize the framebuffer BAT with BAT_G, too. This marks the region
...
as (G)uarded, protecting it from the processor being too aggressive in
reordering and prefetching. This allows OFB_ENABLE_CACHE to work on my
Powerbook G4.
Thanks to matt@ for suggesting this fix. "It's the right thing to do."
2003-02-15 05:07:57 +00:00
b1f5bd8e4f
constify command buffer
2003-02-15 02:36:52 +00:00
9dc8f99b73
Add prototypes for ZS_TXDMA functions.
...
XXX Maybe we should have DMA hooks in MI z8530tty.
2003-02-15 02:28:22 +00:00
baa98646a0
It's not necessary to complicate the computation of `va2pa_offset': we already
...
assume that the few leading pages in front of the kernel image are
always present.
Adjust `va2pa_offset' on sun4 and sun4c as well, so that some day we can
have the bootstrap program load the kernel some place else.
2003-02-14 22:24:58 +00:00
ce385ae9b3
Add CPU IDs for PXA B2 and C0 steppings.
2003-02-14 16:00:33 +00:00
673612c6a0
Use correct segment register in getmementry
2003-02-14 14:39:24 +00:00
1b76f8f55f
One more ServerWorks host bridge, seen on Dell PowerEdge server
...
fix provided in PR port-i386/20333 by Sverre Froyen
2003-02-14 14:26:51 +00:00
136892b8b8
Handle 64-bit daddr_t.
2003-02-14 06:55:17 +00:00
4e7944a9b0
use IPL_SCHED instead of IPL_HIGH to establish interrupt for ACPI;
...
Calling wakeup() from a intr handler established with IPL_HIGH may causes
run/sleep queue to confuse because SCHED_LOCK() uses splsched().
2003-02-14 05:38:39 +00:00
5bcfb63de7
Add PSL_PM
2003-02-14 04:45:32 +00:00
850cd09532
- Change __cpu_simple_lock_t from int to unsigned char.
...
This makes simple lock work on big-endian platforms.
- Assembly-level optimization of __cpu_simple_lock_try().
OK'd by Gregory McGarry. Closes PR port-sh3/20311.
2003-02-14 03:02:41 +00:00
3f573e45ce
Instead of spinning, __asm __volatile("hlt").
2003-02-14 01:11:21 +00:00
ff410ba8d2
If no keyboard is attached, implement RB_HALT with a spin. kern/10481
2003-02-13 21:34:03 +00:00
8268fcff94
Use register 8, not 0, when construct OF_buffer address.
2003-02-13 15:02:49 +00:00
he