Commit Graph

26884 Commits

Author SHA1 Message Date
mrg
9d582d76fc notice siop boot devices. 2001-05-18 23:32:48 +00:00
mrg
2f3c8dae5a correct some comments. <80 char debug statements. 2001-05-18 22:01:57 +00:00
mrg
f42ff7ccae minor cleanups. 2001-05-18 22:01:19 +00:00
mrg
38025bb818 kill dead code. 2001-05-18 21:53:26 +00:00
mrg
e8e471a863 cross compile friendly. 2001-05-18 21:50:45 +00:00
mrg
a9f7a4e194 move the sbus IOMMU work around into the sbus code, so that the psycho
doesn't have to lose a page.
2001-05-18 21:35:23 +00:00
mrg
0bbf8dcb9f update some unused code to reality. 2001-05-18 21:00:47 +00:00
matt
b590763f79 Call __syncicache in db_write_bytes. (makes breakpoints work better). 2001-05-18 20:38:27 +00:00
mrg
0996d006da do not force ebus interrupts to have bit 0x20 set. 2001-05-18 19:17:50 +00:00
fredette
4cb4a99c18 Don't define __LDPGSZ if it's already defined. 2001-05-18 15:34:28 +00:00
fredette
69c174610f Only on the 68010, define a different mcount() function,
one that doesn't use the memory-indirect addressing mode.
2001-05-18 15:33:03 +00:00
fredette
8160155f03 Only on the 68010, assemble versions of copypage
and zeropage that use the 68010's loop mode.
2001-05-18 15:31:38 +00:00
toshii
a90c8b6620 ANSIify and minor KNF. 2001-05-18 14:51:40 +00:00
drochner
9cb9b76e3a cleanup and ELF preparation (_XXX_LABEL(), .align) 2001-05-18 11:51:57 +00:00
enami
6ae30e0e6e Don't touch GIU{USE,TERM}UPDN_REG_W if ONLY_VR4122 is defined.
XXX ... for now.  better way is needed.
2001-05-18 01:41:39 +00:00
matt
78350f3c60 Add ${AOPTS} to AFLAGS 2001-05-17 23:48:31 +00:00
lukem
9c92176b44 don't bother providing duplicate code for {read,write}_eflags() since it's in cpufuncs.h anyway 2001-05-17 16:35:06 +00:00
tsutsui
b9528c4bd8 Call doshutdownhooks() in cpu_reboot().
It was not called when I committed the osiop, and I didn't notice
osiop caused problem when I fixed cpu_reboot()...
2001-05-17 14:53:54 +00:00
enami
c7db455f32 Fix fomula to calcurate TClock; TClock = VTClock / tdiv, not PClock / tdiv.
The manual is inconsistient about this.
2001-05-17 13:10:46 +00:00
enami
c74dc48133 Use VRIP_SIU_ADDR instead of 0x0c000000 for base address of serial console
or kgdb connection.
2001-05-17 06:00:22 +00:00
enami
f2649e3000 Include opt_vr41xx.h and vrcpudef.h 2001-05-17 05:25:32 +00:00
sato
e82d2d52c6 vr4122 releted definition. 2001-05-17 05:04:30 +00:00
chs
8f77918d6c the call to uvm_km_valloc() in pmap_init() can end up triggering a call to
pmap_growkernel() on large memory machines.  at this point we've already
called uvm_page_init(), so we can no longer use uvm_page_physget().
but the flag we use to decide whether to use uvm_page_physget() vs.
uvm_pagealloc() is only set at the end of pmap_init(), so we use
uvm_page_physget() anyway.  switch to using uvm.page_init_done to decide
how to allocate pages, since that's the real indicator of when we have to
change methods.
2001-05-17 02:31:26 +00:00
enami
c7a624dcc6 Allow to pass -d option to kernel. 2001-05-17 01:50:35 +00:00
perry
6859233343 bcopy is dead on i386. the one use left in the kernel was in locore.s
itself, and that got replaced with a call to memcpy. Ross deserves
most of the credit.
note that if you've seen bcopy elsewhere in the kernel, systm.h
defines it as a macro...
2001-05-16 22:09:36 +00:00
scw
c22fb1db55 Finally nobble the bus_dmamap_sync() problem with osiop(4).
Basically, bus_dmamap_sync() `PREREAD' needs to flush the cache
for the start and end of the region if it is not aligned to
a cacheline boundary, otherwise a subsequent POSTREAD can *purge*
valid data which was in the cacheline but *outside* the region
passed to bus_dmamap_sync().

Bus snooping doesn't always help here because osiop(4) calls
bus_dmamap_sync() with POSTREAD even if no data was actually
transferred! (And we can't rely on snooping on the 68060 models anyway).
2001-05-16 19:06:46 +00:00
drochner
35f2b9136a get up-to-date (cleanup, get rid of vm_offset_t) 2001-05-16 18:50:52 +00:00
drochner
2df9b8f13f get up-to-date (cleanup, splimp->splvm) 2001-05-16 18:49:51 +00:00
drochner
038ce6b8af get up-to-date (sprinkle pmap_update()) 2001-05-16 18:49:05 +00:00
drochner
26085708d5 get up-to-date (get rid of vm specific error codes, sprinkle pmap_update()) 2001-05-16 18:48:21 +00:00
drochner
29182d41aa get up-to-date (splimp->splvm, cleanup, sprinkle pmap_update()) 2001-05-16 18:47:04 +00:00
simonb
96a430ba30 If _SOFT_FLOAT is defined, pull floating point va_args from the GPR
area and not the FPR area (by having the double argument type test
always fail).

Fixes problems using <stdarg.h> with -msoft-float.
2001-05-16 15:41:03 +00:00
simonb
9c23ab2d54 Add "switch to ELF". 2001-05-16 15:01:40 +00:00
uch
3c3caa0eb2 don't generate opt_spec_platform.h 2001-05-16 14:26:40 +00:00
msaitoh
cbcd644ce8 sigreturn() fix (from kleink) 2001-05-16 12:42:38 +00:00
enami
43923690d3 Add an entry for Victor MP-C303. 2001-05-16 11:13:20 +00:00
enami
389aaa8b08 Regen. 2001-05-16 11:06:27 +00:00
enami
7cd4cb0d25 Define platid for Victor InterLink MP-C303 2001-05-16 11:04:57 +00:00
tsubai
9b368d61a4 Add AirMac frontend. 2001-05-16 10:56:42 +00:00
enami
fb619e9a70 Calc. cpu speed for VR4122. 2001-05-16 10:49:18 +00:00
leo
cfc3edb550 remove some debugging leftovers. 2001-05-16 08:45:50 +00:00
enami
2f39ebba50 Victor MP-C303 uses 4k pagesize. So, ask system instead of hardcoding it. 2001-05-16 08:40:51 +00:00
enami
422ad4763b The lower bound for getPage is _addr_table_idx, not 0. 2001-05-16 08:37:44 +00:00
enami
679894e9ae Add missing close paren in message. 2001-05-16 08:19:42 +00:00
kanaoka
aa09aa91c7 Search the entire device-space of bus 0 if the router device
address is set to 000:00:0, and the compatible router entry
is undefined.

  Patch PR port-i386/12880 by Dave Sainty <dave@dtsp.co.nz>.
2001-05-16 08:10:36 +00:00
leo
8406481cd3 Regen. 2001-05-16 07:53:54 +00:00
leo
7cab6084a4 Fix errors in previous commit:
- forgot PCI bus
 - forgot pckbd config
2001-05-16 07:50:29 +00:00
matt
c45b32243a Add a VAXANY type. Eventually to used in INSTALL / GENERIC to easily
include support for all VAX platforms
2001-05-16 05:36:53 +00:00
perry
4201bf1ae0 nuke bcopyw and bcopyb. It is unclear what they were for, and nothing
called them.
2001-05-16 05:01:28 +00:00
perry
e266791dab Add a couple of include LINTSTUBs and fix some LINTSTUB prototypes.
Still exits hard because of bcopy/memcpy problem, but that is a
legitimate problem. bcopy should go and be replaced with memcpy to fix this.
2001-05-16 04:18:52 +00:00