summaryrefslogtreecommitdiffstats
path: root/sys
Commit message (Collapse)AuthorAgeFilesLines
* Implement the spirit but not the letter of Terrys hot-char patch.phk1998-02-1318-149/+37
| | | | | | | The differences Terrys patch and this patch are: * Remove a lot of un-needed comments. * Don't put l_hotchar at the front of stuct linesw, there is no need to. * Use the #defines for the hotchar in the SLIP and PPP line disciplines
* Hmm, it is generally an advantage to commit the most recent version ofphk1998-02-133-45/+42
| | | | | ones changes: A faster character painter procedure and fix on cursor bug.
* Use RDMSR instruction instead of WRMSR.kato1998-02-132-32/+32
|
* Sync with sys/i386/isa/npx.c revision 1.56.kato1998-02-131-3/+38
|
* Sync with sys/i386/isa/clock.c revision 1.110.kato1998-02-133-102/+99
|
* Sync with sys/i386/isa/syscons.c and syscons.h revisions 1.254 andkato1998-02-132-73/+106
| | | | 1.37, respectively.
* Staticized.bde1998-02-134-8/+6
|
* Don't (conditionally) override the kernel's min() and max() functionsbde1998-02-132-14/+8
| | | | | | | with macros. This breaks if the functions are replaced by macros with unsuitable semantics. Define a MAX() macro unconditionally instead. max() is unsuitable since we need a constant expression. Don't define MIN() - we never used min().
* Ifdefed SMP-only declarations.bde1998-02-132-6/+16
|
* Don't generate warnings for test coverage checking when compiling LINT.bde1998-02-132-14/+12
| | | | Generate code instead.
* Update timer0_prescaler_count before calling hardclock() while timer0bde1998-02-135-75/+110
| | | | | | | | | | | | | | | | is "acquired". This fixes a TSC biasing error of about 10 msec when pcaudio is active. Update `time' before calling hardclock() when timer0 is being released. This is not known to be important. Added some delays in writertc(). Efficiency is not critical here, unlike in rtcin(), and we already use conservative delays there. Don't touch the hardware when machdep.i8254_freq is being changed but the maximum count wouldn't change. This fixes jitter of up to 10 msec for most small adjustments to machdep.i8254_freq. When the maximum count needs to change, the hardware should be adjusted more carefully.
* Ifdefed some npx code. npx should be optional again.bde1998-02-132-2/+6
|
* Fixed missing privilege checking and off-by-1 bounds checking inbde1998-02-132-12/+18
| | | | | | | | | i386_set_ioperm(). Don't use a magic number for the bound. Fixed missing bounds checking in i386_get_ioperm(). Don't use a magic number for the bound elsewhere in this function. Removed some bogus initializers.
* Ensure that the linker sets for commands exist by putting a standardbde1998-02-133-25/+15
| | | | | command in each of them. This removes the need for hard-to-configure dummy instantiations of the sets.
* Fixed configuration and linkage of ext2_checkoverlap().bde1998-02-134-16/+10
|
* Removed unnecessary dependencies on KERNEL and DIAGNOSTIC. This wasbde1998-02-131-4/+4
| | | | more useful when opt_diagnostic.h had to be included.
* Fixed staticization (the behaviour is undefined when something withbde1998-02-121-1/+1
| | | | external linkage is redeclared as static).
* Unstaticized.bde1998-02-121-2/+2
|
* Add support for VESA mode 0x102 (800x600x4) in syscons. You can activatephk1998-02-126-180/+312
| | | | | | | | | | | | this using option "-b" to the boot blocks. It is smartest to compile a font into your kernel (See LINT), but not mandatory, but apart from the cursor you will see nothing on the screen until you load a font. This mode allows XF86_VGA16 to run in 800x600 mode on otherwise unsupported graphics hardware. A number of buglets in the cursor handling in syscons may become visible this way.
* Fixed initialization of the 4MB page. Kernels larger than about 2.75MBbde1998-02-122-4/+4
| | | | | (from _btext to _end) crashed in pmap_bootstrap(). Smaller kernels worked accidentally.
* Only use the i586-optimized copying and zeroing functions if they arebde1998-02-123-9/+114
| | | | | | actually faster (more than 20% faster for zeroing 1 MB at boot time). This fixes pessimized copying and zeroing on K6's and perhaps on other CPUs that are misclassified as i586's.
* Fix a white-space nit which I happened to notice.phk1998-02-123-9/+6
|
* Add display of type KD_PIXEL.phk1998-02-122-2/+4
|
* Make NO_LKM a new-style option.eivind1998-02-122-2/+4
| | | | Forgotten by: dima
* Oops, sorry, a #define line for my debugging session slipped intoyokota1998-02-123-9/+3
| | | | the last commit ;-<
* Sync with sys/i386/isa/syscons.c revision 1.251.kato1998-02-121-129/+183
|
* Sync with sys/i386/conf/options.i386 revision 1.72.kato1998-02-122-2/+6
|
* Replace non-existent ip_forwarding with ipforwardingache1998-02-121-2/+2
| | | | (compilation error)
* Alter ipfw's behavior with respect to fragmented packets when the packetalex1998-02-121-3/+35
| | | | | | | | | | | | | | | | | | | offset is non-zero: - Do not match fragmented packets if the rule specifies a port or TCP flags - Match fragmented packets if the rule does not specify a port and TCP flags Since ipfw cannot examine port numbers or TCP flags for such packets, it is now illegal to specify the 'frag' option with either ports or tcpflags. Both kernel and ipfw userland utility will reject rules containing a combination of these options. BEWARE: packets that were previously passed may now be rejected, and vice versa. Reviewed by: Archie Cobbs <archie@whistle.com>
* I'm not sure whether this is a correct way to do it,dima1998-02-114-6/+18
| | | | | | but here's a new kernel option - "NO_LKM" If anyone has better ideas - please let me know.
* Fix a && that should be an &.dg1998-02-111-2/+2
| | | | | Reviewed by: "John S. Dyson" <dyson@FreeBSD.ORG> Submitted by: jwd@unx.sas.com (John W. DeBoskey)
* Only forward source routed packets when ip_forwarding is set to 1.guido1998-02-111-2/+2
| | | | | | | | | This means that a FreeBSD will only forward source routed packets when both net.inet.ip.forwarding and net.inet.ip.sourceroute are set to 1. You can hit me now ;-) Submitted by: Thomas Ptacek
* Remove description on the flags for psm. They are adequatelyyokota1998-02-113-30/+3
| | | | documented in the man page for psm(4).
* Added SC_DISABLE_REBOOT and SC_MOUSE_CHAR.yokota1998-02-112-2/+6
|
* Clean up comments describing Luigi's alternate sound driver.jkh1998-02-113-12/+27
|
* - Move the mouse cursor to the center of the screen after videoyokota1998-02-113-48/+123
| | | | | | | | | | | | | | | | | | | | | | | | | mode switch in ioctl. Possibly related to PR: kern/4271 - A kludge: initialize scp->xpixel and ypixel even in the text mode. If the console enters the `unknown' graphics mode via the ioctl KDSETMODE (KD_GRAPHICS), these fields are not set (because syscons cannot know the correct values), but set_mouse_pos() need to refer to these field to adjust the mouse position. - Turn off MOUSE_VISIBLE when switching video mode by ioctl. - another new option: SC_MOUSE_CHAR Define the first character code of four consecutive codes to be used for the mouse cursor. Default codes are 0xd0 through 0xd3. Beware that if you decide to use any codes outside the range of 0xc0-0xdf, the mouse cursor may not look good, because of the way VGA displays characters in 9-dot-wide character cells. Requested by several people. (This patch was tested by a person who recently reported, in the -current ML, a page fault problem in the kernel (draw_mouse_iamge()) after X server shutdown. The patch cured his problem.)
* - A new semaphore: font_loading_in_progress.yokota1998-02-113-27/+81
| | | | | | | | | | | | Don't touch/update the screen while manipulating font data. Possibly related to PR: kern/4271 - Set up VGA in alphanumeric mode rather than graphics mode when loading font into video memory. This will drastically reduce flicker. PR: bin/2977 - Set up scp->font_size properly during video mode switch caused by ioctl.
* - Break scrn_timer() into two pieces. Now screen update is done in ayokota1998-02-113-255/+291
| | | | | | | | | | | separate routine: scupdate() called from scrn_timer(). - Make sure that the screen is updated for the low-level console routines sccngetc() and sccncheckc(). A new routine, sccnupdate(), is introduced and will call scupdate() above. Requested by: bde and msmith OKed by: sos
* - Ignore KBD_RAW_MODE and KBD_CODE_MODE in scgetc() when it's calledyokota1998-02-113-27/+54
| | | | | | | | | | | | | | | | | | | | | | from the low-level console routines sccngetc() and sccncheckc(). Submitted by: bde (a long time ago) - Don't try to ring bell and immediately return from do_bell() while device probe is in progress at boot time; the timeout queue is not functional yet. PR: kern/2424 - Stop running the screen saver after panic() is called: check if `panicstr' is non-NULL during scrn_timer(). PR: kern/5314 - A new option: SC_DISABLE_REBOOT The reboot key (usually Ctl-Alt-Del) will be ignored if this option is defined. You may still have the reboot key defined in the keymap and it won't cause error when the keymap is loaded, but it will be quietly treated as nop. OKed by: sos
* De-staticize enough to make all the LKMs work again. Add comments whereeivind1998-02-114-10/+10
| | | | deemed relevant.
* Include SIMPLELOCK_DEBUG functions even if SMP if compiling LINT; giveeivind1998-02-111-2/+6
| | | | an error for the combination if _not_ compiling LINT.
* Back out statification and restore a declaration that got lost.eivind1998-02-102-4/+8
|
* Staticize. (Diffs by me & and Simon in cooperation.)eivind1998-02-104-43/+43
|
* Fix warning after previous staticization.eivind1998-02-102-4/+4
|
* Move include of <machine/ipl.h> inside ifndef SMP where it is used, toeivind1998-02-101-3/+3
| | | | avoid getting 'unused include file' warnings in the SMP case.
* Deleted unused variable.kato1998-02-102-4/+2
|
* Sync with sys/i386/i386/userconfig.c revision 1.102.kato1998-02-101-2/+15
|
* Undo UN_KLOCK hack except union_allocvp(). Now, vput() doesn't lockkato1998-02-106-108/+64
| | | | the vnode.
* Added entry of 16-bits bus lnc card.kato1998-02-102-4/+4
| | | | Submitted by: Chiharu Shibata <chi@rd.njk.co.jp>
* Support C-NET(98)S ethernet card (PC-98 only).kato1998-02-104-0/+210
| | | | Submitted by: Chiharu Shibata <chi@rd.njk.co.jp>
OpenPOWER on IntegriCloud