Commit Graph

874 Commits

Author SHA1 Message Date
pk 90fb0cc9da Don't come down so hard on targets that ignore SDTR msgs. 1996-03-05 09:29:58 +00:00
chuck 9bb0e5dba5 Fix two bugs reported by Don Koch <aardvark@poirot.krl.com> (PR#2169)
[these bugs apply to the xy as well as the xd driver]
 [1] check return value from malloc() for NULL before trying to bzero it.
 [2] use "=" rather than "|=" when writing to CSR (otherwise you may
        ACK something you don't mean to!).
1996-03-04 20:13:39 +00:00
chuck 9fee13d104 Fix two bugs reported by Don Koch <aardvark@poirot.krl.com> (PR#2169).
[1] check return value from malloc() for NULL before trying to bzero it.
 [2] use "=" rather than "|=" when writing to CSR (otherwise you may
	ACK something you don't mean to!).
1996-03-04 20:11:32 +00:00
cgd 18ec26aa21 add _MACHINE and _MACHINE_ARCH, which are like MACHINE and MACHINE_ARCH,
execpt without quotes.  meant to be __CONCAT()ted for easy #includes
of machine-dependent headers for MI code (e.g. for the MI ISA/EISA/PCI/TC
bus code).
1996-03-04 05:04:10 +00:00
pk 855b6def84 Add ns_cksum. 1996-03-02 23:55:17 +00:00
thorpej 9f2837ac9b "oops", forgot to check this in...add entries for the cgfour and cgeight. 1996-03-01 08:11:00 +00:00
thorpej f579d1a485 Fix an operator precedence botch. *sigh* This just hasn't been my week. 1996-03-01 07:44:43 +00:00
pk abd3cf4df2 Don't reset `xs->retries'. 1996-02-29 23:33:48 +00:00
pk bf59310f69 Fix typo, and bring back VA hole handling. 1996-02-29 22:15:13 +00:00
cgd 3d67685d36 Do not build vnode_if.[ch] for each kernel. Build them once, like the
various syscall sources/headers, and just compile them.  From PR 2142, OK'd
by mycroft.
1996-02-29 20:54:58 +00:00
thorpej 93dff23036 Back out the "don't use DMA if SCSI_POLL is set" saftey-net. The
MI 5380 code did something I didn't really expect, which caused DMA
to disable itself if only DMA was enabled.  *grumble*
1996-02-29 03:28:38 +00:00
gwr 2de7b2e5cb update PMAP_PREFER 1996-02-28 22:44:33 +00:00
pk 8bea3309c3 Fix typo 1996-02-28 22:09:27 +00:00
thorpej 50c8d6cdb0 Oops...missed to places where I could have used CPU type macros. 1996-02-28 20:53:02 +00:00
thorpej edf9d66f32 First-cut support for P4 framebuffers. Currently supports P4 variants
of bwtwo, cgfour, cgsix, and cgeight.  Includes support for attaching
the overlay plane of the cgfour and cgeight as a bwtwo instance (a'la
SunOS).

P4 register descriptions, cgfour driver, and cgeight driver partially
derived from OpenBSD.

This code may still need polishing.
1996-02-27 22:09:14 +00:00
pk fe5a56ec9d Be sure to malloc() in page-sized units. 1996-02-27 13:17:01 +00:00
pk 70eee90be3 remove a printf 1996-02-27 09:10:24 +00:00
pk 9216cf3276 Drain rev 1 chips in dma_reset(). 1996-02-27 00:36:11 +00:00
pk 2c646f8da5 `bwtworeg' is just about the same as the generic `fbcontrol'. 1996-02-27 00:32:34 +00:00
pk 6b5eb65b21 bt_reg => fbcontrol. 1996-02-27 00:14:17 +00:00
pk 8628f94179 Use `video enable' bit in framebuffer control. 1996-02-27 00:11:14 +00:00
pk 2fe54615fb Define a couple more framebuffer control bits. 1996-02-27 00:09:23 +00:00
pk 517988b940 Interims:
Comment out `STATINT' checking; this is apparently unreliable on some
ESP revisions.
Deal with target initiated synch re-negotiation.
1996-02-26 14:48:30 +00:00
mycroft 2ce52ae489 Define build programs with `?=', so they can be overridden with `makeoptions'. 1996-02-26 02:53:20 +00:00
pk a3ab0e81da Use CPU-type macros. 1996-02-25 22:03:20 +00:00
pk 3d182fe265 Use CPU-type macros 1996-02-25 21:53:49 +00:00
pk ba0eb7ea59 Use CPU-type macros. 1996-02-25 21:45:53 +00:00
thorpej 1a5d5e25c6 Add an "sw_options" patchable global so that folks can still live life
on the edge (of sw DMA uncertainty) with pre-built kernels.
1996-02-23 16:29:34 +00:00
thorpej ef3500b579 Attempt to work around a hardware bug that screws up DMA on the 4/100.
Apparently, some early 4/100 DMA controllers do illegal memory access on
large ( >= NBPG ) transfers at the end of the transfer.  This appears
as SI_CSR_DMA_BUS_ERR in the csr.  To work around this, we simply
transfer the (up to 3) missing bytes from the bpr.  We were doing this
anyway, so the work-around is to ignore the bus error.

BUT!  I goofed when I implemented the "left-over byte" code for the sw!
It *should* be correct now.  Keep metrics (acceeible via DDB) on the number
of 1, 2, and 3 byte clean-ups, as well as the number of "clean" transfers,
just so we can get a clearer picture.

Thanks to Andrew Gillham <gillham@whirlpool.com> for noticing this!
1996-02-23 07:24:45 +00:00
mycroft 291267e40c Implement SDEV_AUTOSAVE. 1996-02-22 23:35:04 +00:00
pk 2f3ad94f8d An #ifndef _LOCORE to make this file safe to include in locore.s.
Introduce a couple of `CPU-type' helper macros to reduce the preprocessor
tangos in many a source file.
1996-02-22 21:59:08 +00:00
pk e259a02425 Correct some prototypes. 1996-02-22 15:04:07 +00:00
thorpej c86505a6a5 If we get a transfer with the SCSI_POLL flag set, don't use DVMA. We
only get these during autoconfiguration and during crash dumps.  During
autoconfiguration, the transfers are small enough that DVMA won't be used
anyway.  However, using DVMA during a crash dump can be dangerous,
depending on the nature of the panic, so we avoid it.
1996-02-22 07:25:05 +00:00
thorpej 2f132eb5e7 Bring in Gordon's fix from the sun3 port:
Correct the DMA transfer count when the target disconnects before
	the whole transfer is completed.  (Affects VME writes)
	Reselect now works on the VME si board!
1996-02-22 07:14:53 +00:00
thorpej a246f6c835 Fixup some RCS ids. 1996-02-22 06:50:49 +00:00
pk 66fedbbe19 Correct format in a printf(). 1996-02-21 10:21:14 +00:00
chuck adeed1d823 minor clean up: revise locations of disk_busy/disk_unbusy calls to match xd.c 1996-02-21 03:36:51 +00:00
pk b9de283ae6 Fix typo; pointed out by jtk. 1996-02-19 10:30:09 +00:00
pk 9b46365465 Adapt to changed kernel data structures (PR#2100). 1996-02-19 09:46:32 +00:00
thorpej a9012450ec Fix bug in bwtwounblank() ... didn't have code to enable video on sun4/200
obio framebuffer.  Noticed when my 4/260 dropped into DDB and the screen
didn't unblank.  Pull all of the video enable/disable into functions so
this mishap doesn't happen again.
1996-02-19 00:15:46 +00:00
pk a5ea7547d2 Correct timetochip() prototype.
Cleanup microtime() a bit.
1996-02-18 15:38:41 +00:00
pk 1a62c0b31b delay() takes an unsigned int; get the code to match. 1996-02-17 09:47:48 +00:00
pk 1caa80a432 Bring prototypes into scope.
Fix a few bogus `return's.
1996-02-17 00:03:17 +00:00
pk f1518de67c delay() prototype. 1996-02-16 22:13:34 +00:00
pk cdb37b19d3 Account for the fact that `clockreg' might not be page-aligned.
Misc prototype goo.
1996-02-16 22:12:13 +00:00
pk da1402dcb7 Eliminate <setjmp.h>. 1996-02-14 01:03:17 +00:00
pk fe9d17199a If `cold' simply halt the machine in boot(). Note, this still leaves a window
where vfs_shutdown() and sync() can be called prematurely.
1996-02-13 23:59:02 +00:00
pk 54bd5c6954 Fix ANSI prototype goof (from Openbsd).
Add explicit return type to most functions.
V3 PROM console support.
1996-02-13 22:49:48 +00:00
pk 773f74b60d Change some caddr_t's to void *'s. 1996-02-13 22:43:33 +00:00
pk 58cd67c20d Make sure we assign a value to variables we use.
Don't express size of NVRAM in terms of machine page size.
1996-02-13 22:38:25 +00:00