src - OpenBSD base system

Age	Commit message (Collapse)	Author
2003-05-13	remove -fno-stack-protector (enable propolice)	Ted Unangst
	ok deraadt@
2003-05-13	The current solution to handle the protection fault trap is not	Artur Grabowski
	correct. It breaks down if we're trying to jump through a function pointer. The protection fault trap on i386 must be one of the most braindead traps ever invented in the history of humankind. It doesn't give you any information about what went wrong except the instruction that faulted. Since the problem we're trying to deal with is a segmentation problem, we don't get the desitination that we want to jump to, we just get the instruction and we won't add a disassembler to trap handling just to try to figure out what went wrong. What we want to do is to handle this as a normal fault to let noexec accounting in pmap_enter deal with the changes to the code segment. Unfortunately that's impossible. We don't know the faulting address, so we need to change how the exec accounting works. Basically the code segment must already cover the address we want to execute before we can fault it in. New scheme: o Start with conservative code segment. o If we get a protection fault, go through all mappings in the process and find the highest executable mapping, fix up the code segment and record that address. If the code segment didn't change, the protection fault wasn't fixable - just die. o If the highest executable mapping is removed, just reset the code segment to something conservative and let the next protection fault deal with it. We can't read all the vm mappings of the process from the pmap because of locking hell. This should allow floating code segment whenever someone implements that. Also, fix the pmap_protect function to behave more like the other pmaps we have and be slightly more agressive to force more proper protection changes. ok:ed by various people.
2003-05-13	kill a stupid warning	Jason Wright

2003-05-13	nuke curproc common (yes this will need revisiting in smp land, but what	Jason Wright
	doesn't).
2003-05-12	use new display types; ok miod	Jason Wright

2003-05-12	use new display types	Jason Wright

2003-05-12	Initial work to bring this driver in par with other sparc frame buffers:	Miod Vallat
	- do not reprogram the colormap to black and white, after programming the rasops one (oops). This gives us back a black on white display with colour support. - if the screen resolution would cause the switch from the prom font to a reasonably sized wsfont to leave text in the margins, clear the screen (cut and paste from similar code in vigra.c) - try and use more symbolic constants rather than magical values all over the place. Thanks to millert@ for bringing a sparcbook here.
2003-05-12	Enable Intrepid's mac-io match.	Thierry Deval
	Ok drahn@
2003-05-12	kill another common; drahn ok	Jason Wright

2003-05-12	add a few more trap types	Jason Wright

2003-05-11	nuke another common: want_resched	Jason Wright

2003-05-11	nuke common: physmem	Jason Wright

2003-05-11	don't run off end of strings and use bad pointers	Marcus Watts
	iterate boot list at most once allow kernel names with - in them complain about option strings that do not start with - distinguish between short read & bad read don't quit if a file can be opened but can't be booted
2003-05-11	string cleaning; krw ok	Theo de Raadt

2003-05-11	nuke debugging stuff	Jason Wright

2003-05-11	get the offset of the config space correct, now to figure out why the rtc is ↵	Jason Wright
	unhappy.
2003-05-11	check done for NULL before dereferencing. ok drahn@	Ted Unangst

2003-05-10	string cleaning; ok beck dhartmei	Theo de Raadt

2003-05-10	Remove some exec mappings in the pmap test code (ifdef DEBUG).	Artur Grabowski

2003-05-09	Map page tables non-exec.	Artur Grabowski
	deraadt@ ok.
2003-05-09	No need to map the message buffer executable.	Artur Grabowski
	fries@ niklas@ ok
2003-05-09	No need to map framebuffers executable. Really.	Artur Grabowski
	fries@ niklas@ ok.
2003-05-09	'found' variable was not used correctly, simplify the logic.	Dale Rahn

2003-05-09	sprintf() removal for ppc disassembler. ok millert@	Dale Rahn

2003-05-09	Remove dead (ifdef 0) code that pollutes my greps.	Artur Grabowski

2003-05-08	'branch absolute linking' instruction is 'bla', not 'bal'.	Dale Rahn

2003-05-08	KNF	Dale Rahn

2003-05-08	Change extract_field() arguments, eliminates the nasty '31 -' all over the	Dale Rahn
	file. Remove NetBSD tag, this file is not used by NetBSD. Copyright cleanup.
2003-05-08	Put newline on end of disasm so that x /i <addr>,<cnt> works correctly.	Dale Rahn

2003-05-08	EXAMPLE -> EXAMPLES in .Sh section, as per mdoc template.	Jason McIntyre
	crunch stuff still needs a bit of work.
2003-05-07	string cleaning	Theo de Raadt

2003-05-07	double semicolon in local var decl	Michael Shalayeff

2003-05-07	disable profiling for tlb fault handlers	Michael Shalayeff

2003-05-07	selecet pa-level and scheduling separately	Michael Shalayeff

2003-05-07	double memsize as there might be prom as well	Michael Shalayeff

2003-05-07	exit stack is not used anymore. fpemu and emergency are fine w/ just a page ↵	Michael Shalayeff
	per face
2003-05-06	allow programming dma block bus transaction length on 710 and set to 8 words ↵	Michael Shalayeff
	for hppa/osiop; krw@ ok
2003-05-06	Declare the major number for wd, so that a custom kernel can specific explicit	Miod Vallat
	root on wd0a. Problem spotted by Denis A. Doroshenko.
2003-05-05	Move exec base to 0x1c000000, exe/data gap to 512MB. Allows better	Dale Rahn
	interleave of exe/shared libs. Raise MAXDSIZ back to 1G. This change REQUIRES a binary update on i386.
2003-05-05	Write explicit scsibus at controller attachment, rather than scsibus at scsi	Miod Vallat
	attribute.
2003-05-04	mop-mop squeek-squeek	Michael Shalayeff

2003-05-04	snprintf	Theo de Raadt

2003-05-04	Set code segment register according to nxpages, copied from machdep.c	Dale Rahn

2003-05-04	When a protection fault occurs, force %cr2 to contain the pc as	Dale Rahn
	the fault address.
2003-05-04	string cleaning. ok deraadt@	Ted Unangst

2003-05-02	when flipping the code descriptors also update cs in the	Michael Shalayeff
	tss and not only in the frame since we might be returning that way too. add a heuristic for detecting an exec protection fault: iff we get a read protection fault (which we normally never get due to our segments being always readable) we assume that it was an exec protection indeed and go to page fault routine which will decide the rest for us (including sending a signal should that be needed). problem found by drahn@ and testing by many ppl.
2003-05-02	add eap and do a bit of cleaning	Jason Wright

2003-05-01	add eso, opl, and midi, and clean the audio stuff up a bit	Jason Wright

2003-04-30	do not be hungry for buffers and allocate lower than a max (for iodc io)	Michael Shalayeff

2003-04-30	from netbsd:	Michael Shalayeff
	The AMD-K5 Model 0 gets the PGE bit in the CPU features word wrong using the APIC bit instead, according to the AMD Processor Recognition App. Note. Add a fixup routine to patch up cpu_feature in this case. XXX Need a way to kick the pmap to enable pmap_pg_g -- look at how the MP branch deals with this.