Commit Graph

262 Commits

Author SHA1 Message Date
matt ff5f5ea027 Don't compile SPL/INTR stuff if NEWINTR is defined. 2001-09-05 17:08:41 +00:00
matt 779b9b4649 Fix some register definitions. 2001-09-05 17:05:36 +00:00
matt 0b5b9ca90e Change <machine/irqhandler.h> to <machine/intr.h>. Change
{irq,fiq}handler_t to struct XXXhandler
2001-09-05 16:23:10 +00:00
matt 792b7ebd13 Change <machine/irqhandler.h> to <machine/intr.h> since the latter always
includes the former and is a standard include file.
2001-09-05 16:17:35 +00:00
matt c9fd88afcf Don't clean the minidata cache by default on XScale. 2001-09-05 16:14:49 +00:00
matt 4370f884b2 Flesh XSCALE support. 2001-09-05 16:13:18 +00:00
matt 78944c67e3 Add an obio bus. 2001-09-05 16:12:36 +00:00
wiz b940fa8a83 Fix typos (CONS instead of CONF, and specifc instead of specific). 2001-09-04 21:45:19 +00:00
matt 7ce6450eac Add a merged <pcb.h> for arm26/arm32. 2001-09-03 19:20:27 +00:00
matt d7c7a3ea02 Add __ELF__ stuff for resized types under ELF. 2001-09-03 01:51:39 +00:00
matt b0f89169ee Switch netwinder to use a common conf.c for the arm ports. 2001-09-03 01:50:00 +00:00
simonb 023bdd2bf2 Use comments around the token after a #endif. 2001-08-31 04:44:54 +00:00
matt e4892d7157 When invaliding single icache line on XScale, the branch target buffer
also needs to be invalidated.  Also, but in the right sequence to empty
the pipeline on XScale.
2001-08-30 01:15:39 +00:00
chris 37fb92dcab Remove unused entries from the pcb. This is with a long term view of merging the arm26 and arm32 pcb's. 2001-08-27 11:39:43 +00:00
matt edc83ba7f4 Add xscale cpufuncs (using the sa110 ones when they are identical). 2001-08-27 00:00:27 +00:00
matt e8f317bba8 Add xscale cpu functions 2001-08-26 19:41:57 +00:00
matt 4e642cc5fd Add i80312 register definitions (just registers for now). 2001-08-26 19:25:47 +00:00
matt 08f4bcddf3 files file for xscale (i80312 aka iopxs) 2001-08-26 19:19:09 +00:00
bjh21 ed5e27cc44 Increase the size of a jmp_buf when using ELF, since this will be our last
chance to do so for a while.  We don't use the extra space yet, but VFP
support is likely to need it.
2001-08-25 14:45:59 +00:00
chris 140252be2b Arm has a vac, so we must use pmap_enter/remove for vmapbuf rather than k* versions, otherwise we may not be doing the right caching thing. 2001-08-20 21:52:09 +00:00
wiz d375535396 Fix typo in comment (suceed). 2001-08-20 11:49:11 +00:00
matt 4bc8190b62 Change local offset/label for irqhandlers to Lirqhandlers. Because in
ELF _C_LABEL(irqhandlers) == irqhandlers so we were getting self-referential
pointer which causes the IRQ to fall on itself.  [With this change the
netwinder kernel can now be *ELF* and boots to multiuser]
2001-08-19 07:29:26 +00:00
jdolecek c495131fe3 Move _insque()/_remque() to libkern. Once remaining uses would
be converted to <sys/queue.h> macros, _insque()/_remque() would be eliminated
altogether.
2001-08-12 08:35:31 +00:00
chris 0b0518e0dd Fix compile without DIAGNOSTICs enabled 2001-08-11 14:47:56 +00:00
chris 191de4d242 Checking a whole host of pmap changes:
Improved locking (not that we actually use it on a uniprocessor, but one day :)
Removed unneeded splvm's
tweaked pmap_clean_page code to only flush the cache if the page is mapped in the current pmap  (based on diff from richard E)
Adopted pv entry allocation mechanism from i386.
Laid framework for returning ptp's when we've finished with them rather than holding onto them till the process exits.
ptp's are now allocated with a uvm object for the pmap, means that we can walk a list to free them off in pmap_release, until they get freed off by pmap_remove.

Also implemented a page zeroing function when the processor is idling.  Note that hpcarm may wish to disable this.

I believe this code to be stable, if anyone has any problems please shout up.
2001-08-11 13:20:19 +00:00
chris c5de587666 Make use of pmap_kenter instead of pmap_enter for kernel buffers. 2001-08-11 13:00:48 +00:00
chris 163b9983d7 Add arm_trunc_page, similair to trunc_page but using a hardcoded PGOFSET. 2001-08-11 12:59:45 +00:00
chris 6cca1c3f58 Update to make use of a proper kenter implementation for vmapbuf and vunmapbuf. 2001-08-11 12:57:25 +00:00
chris 6e43d02452 Patch in Richard E's changes to bcopy page, the loop doesn't need unrolling as much as it currently is due to the overhead of the load/stores taking most of the time. (profiling shows minimal time difference) 2001-08-11 12:44:42 +00:00
bjh21 4dd0161267 Use STACKALIGN rather than &=~7. 2001-08-07 22:56:09 +00:00
bjh21 8fa03d16ac Add macros to hide the details of stack-pointer alignment. 2001-08-07 22:54:00 +00:00
bjh21 bd6d4d6b77 Use <arm/param.h> to get shared definitions. 2001-08-07 22:26:56 +00:00
bjh21 f37b49755f When delivering a signal, ensure that the stack is 8-byte-aligned. 2001-08-06 22:32:23 +00:00
chris 9522cf08d4 Share procfs_machdep.c across arm. 2001-08-05 13:09:45 +00:00
matt 9509e734a3 Force size_t formats/arge to be (u_long). I'd use 'z' for this but gcc
2.95.3
2001-08-05 05:07:27 +00:00
rjs faf8b4c913 Update from hpcarm with changes to probe routine and test for chip ID. 2001-08-01 14:02:58 +00:00
matt 03b77c0534 Add a CPU_XSCALE defopt (for eventual support). 2001-07-31 17:00:24 +00:00
rjs 6b65fea380 Remove unused hpc includes. 2001-07-30 16:23:09 +00:00
rjs 9013c85405 Update from hpcarm to remove iPAQ egpio code. 2001-07-30 16:20:06 +00:00
rjs 51bc7258f8 Update from hpcarm with MCP base definition. 2001-07-30 16:08:42 +00:00
rjs c79059dd3f Update from hpcarm to remove iPAQ and Jornada specific #defines. 2001-07-30 15:58:56 +00:00
rjs b85ae98e65 Update from hpcarm with MCP register definition. 2001-07-30 12:19:04 +00:00
rjs 495d108bd3 Update from hpcarm with cookie change. 2001-07-30 10:55:49 +00:00
rjs 9b65c24aa5 Remove unused LCD definitions. 2001-07-30 10:50:59 +00:00
rjs 1304d2220e Mirror removal from hpcarm. 2001-07-30 10:48:31 +00:00
rjs 499d1068e8 Update from hpcarm. 2001-07-30 10:46:08 +00:00
rjs e35b542831 Mirror removal from hpcarm. 2001-07-30 10:33:59 +00:00
rjs 730b1a57ac Copy from hpcarm. 2001-07-30 10:17:21 +00:00
chris fa4e4d9668 Rework the pmap_release code to not have to walk the ptpt, it now uses a uvm_object to track the allocated vm_pages, this means it can free off the entries in the uvm_object.
Testing shows that it's about 5% faster on the make configure step for gmake.
2001-07-29 12:45:27 +00:00
chris 89415a32e2 PR 13581
Add empty pmap.h and vmparam.h to make userland happier.
2001-07-29 09:50:07 +00:00