src - OpenBSD base system

Age	Commit message (Collapse)	Author
2013-05-21	Small code clean up. Removing noisy printfs.	Brandon Mercer

2013-05-21	Check for NULL ptr before dereferencing, not after; found the hard way by	Miod Vallat
	dcoppa@
2013-05-21	Remove some unused defines and make the number of interrupt lines variable.	rapha
	ok bmercer@
2013-05-21	Ensure that sc_lock is held when calling sr_meta_native_bootprobe() - the	Joel Sing
	softraid status functions will complain otherwise. Found the hard way by ajacoutot@
2013-05-21	sr_raid_startwu() is no longer called from interrupt context, hence does	Joel Sing
	not need to invoke a workq-based callback.
2013-05-21	Provide a function that handles the scheduling of work units. This	Joel Sing
	simplifies the discipline code, avoids code duplication and moves the scheduling logic into a single location. ok krw@
2013-05-21	Add a driver for the am335x timers, to be used by the beaglebone.	rapha
	Derived from the existing omap3 gptimer. ok patrick@
2013-05-21	Make the sr_raid_startwu() function handle deferred work units, instead of	Joel Sing
	the caller having to handle the state changes.
2013-05-21	Fix missing work unit state initialisations.	Joel Sing

2013-05-21	Use a state to indicate that a work unit should only be constructed and not	Joel Sing
	scheduled, rather than trying to imply this from the rebuild flag.
2013-05-20	de-static-ize, such that people don't get confused by aesni failures on	Mark Kettenis
	AMD hardware the next time somebody tries to run on a broken KVM hypervisor.
2013-05-20	Remove `abort_task' from usb task queue before recycling a `struct	YASUOKA Masahiko
	usbd_xfer object' which includes the `abort_task'. Otherwise usb_abort_task_thread() may try to dequeue the recycled task then it causes panic with page fault. reported by Edd Barrett and Wade, Daniel. ok mpi
2013-05-19	Switch mvme68k to MI wdsc(4) driver. Tested on MVME147.	Miod Vallat

2013-05-19	select(-1, ...) should fail with EINVAL.	Philip Guenther
	ok matthew@
2013-05-18	fix cut'n'paste abuse in "can't happen" error message	Miod Vallat

2013-05-18	Add parameters describing the usable part of the GTT and enable the checks	Mark Kettenis
	that use them.
2013-05-18	Get rid of the pmap7 header now that we merged it into the pmap one.	Patrick Wildt

2013-05-18	Bring beagle back to the other pmap header now that it's merged.	Patrick Wildt

2013-05-18	Modify pmap to work with the pmap header used on armv7. Merge both	Patrick Wildt
	headers so that we only need one of them. "Go for it." miod@
2013-05-18	Mask out the TEX remap and the Access Flag bits when setting the	Patrick Wildt
	system control register. Also actually use the mask. ok miod@
2013-05-18	Make sure we executed the instruction before continuing. Also	Patrick Wildt
	replace calls to drain the write buffer with the correct ones for armv7. ok miod@
2013-05-18	Use the actual armv7 dcache_inv_range function instead of the wbinv one.	Patrick Wildt
	ok miod@
2013-05-18	Sync with Linux. Mostly comment and whitespace changes, but adds some missing	Mark Kettenis
	locking, and enables some #ifdef'ed out code.
2013-05-18	Missing bits intended to go with the previous commit but lost in diff splitting:	Miod Vallat
	Split inittodr/resettodr code from actual tod routines, to allow for multiple tod routines in the near future. Use the mi clock_ymdhms_to_secs and clock_secs_to_ymdhms routines.
2013-05-18	enw->new in comment;	Jason McIntyre
	spotted by krw, missed by jsg, and forwarded From: Alexey Suslikov
2013-05-17	Fix 'the right-of-cursor background color is inverted when we do	Kenji Aoyama
	delete-after-cursor' bug on luna88k wscons console, by adding the same operation in om_putchar() to om_erasecols() and om_eraserows(). ok miod@
2013-05-17	Preliminary support for MVME180 and MVME181 boards. On-board serial ports	Miod Vallat
	and VME controller are supported; tod chip isn't supported yet (coming soon). MVME236 memory boards are not supported and won't likely be (unless I get my hands on one). MVME181-1 boots multiuser (slowly) with MVME328 SCSI and MVME376 Ethernet despite only having 8MB of memory onboard. MVME180 untested; might need tweaking to get MVME SCSI and Ethernet boards supported. Many thanks to Matti Nummi for lending me an MVME181 board.
2013-05-17	Replace the bunch of md_* function pointers with a `struct board' containing	Miod Vallat
	function pointers for all the board-specific code. Add a bunch of `struct board' methods to cover most, if not all, of the `per-board' logic. This allows most of the md drivers to be cleaned up and no longer need to embed board-specific knowledge.
2013-05-17	Move interrupt-related defines from <machine/cpu.h> to <machine/intr.h>.	Miod Vallat
	No functional change.
2013-05-17	Split inittodr/resettodr code from actual tod routines, to allow for	Miod Vallat
	multiple tod routines in the near future. Use the mi clock_ymdhms_to_secs and clock_secs_to_ymdhms routines.
2013-05-17	Unnecessary include	Miod Vallat

2013-05-17	Extend cmmu routines to return the caching mode to use for page tables.	Miod Vallat
	Alter the 88200-specific code to enforce cache-inhibited page tables for extremely old 88200 versions, and to disable write-back caching on systems where xmem instructions do not behave correctly when applied to write-back cached addresses. No change introduced on 88110 systems, as well as most 88100 systems; the affected systems are 88100 systems with 88100 revision < 10 and/or 88200 revision < 7; that is, only early MVME181 and MVME188 (not 188A) systems.
2013-05-17	Fix comments telling that the `serial mode' bit in PSR only applies to 88110,	Miod Vallat
	it also exists on 88100. While there, do not allow userland to set it.
2013-05-17	Add a dummy IPL_MPSAFE definition.	Mark Kettenis
	ok miod@, mikeb@
2013-05-17	Implement IPL_MPSAFE for pci bus interrupt handlers.	Mark Kettenis

2013-05-17	wsmoused support	Mark Kettenis

2013-05-17	Implement wsmoused support based on the new multiple screen support. Only	Mark Kettenis
	works if the RI_VCONS flag is specified when initializing a rasops_info descriptor.
2013-05-17	Do not attempt to delete the undeletable RNF_ROOT route.	Kenneth R Westerback
	Discovered & fix tested by Peter J Philip. ok claudio@ blambert@
2013-05-17	Make it clear that the code related to a transfer submission doesn't	Martin Pieuchot
	leak anything to userland because it doesn't set the USBD_SHORT_XFER_OK flag. Also prevent a bad copy/paste from introducing a similar issue by using the actual transferred length instead of the requested one in uiomove(). ok miod@
2013-05-17	Don't leak information to userland in case the actual transfer length is	Martin Pieuchot
	smaller than the requested one. From ws@NetBSD via miod@. This problem can only occur when the USBD_SHORT_XFER_OK flag is set, otherwise completed transfers with a length smaller than the one submitted are treated as errors. ok miod@
2013-05-17	Move an extern declaration into its corresponding header file.	Martin Pieuchot

2013-05-16	burner support	Mark Kettenis

2013-05-16	Implement a mechanism to establish interrupt handlers that don't grab the	Mark Kettenis
	kernel lock upon entry through a new IPL_MPSAFE flag/level.
2013-05-16	Implement a mechanism to establish interrupt handlers that don't grab the	Mark Kettenis
	kernel lock upon entry through a new IPL_MPSAFE flag/level. ok mikeb@, mpi@
2013-05-16	Ignore colormap operations at 1bpp, to work with Xorg wsfb driver.	Kenji Aoyama
	Still need more work in Xorg... ok miod@
2013-05-16	attach to new RTL8188RU id, from Sigi Rudzio	Stuart Henderson

2013-05-16	sync	Stuart Henderson

2013-05-16	add another RTL8188 id, from Sigi Rudzio	Stuart Henderson

2013-05-16	Add missing IPL_AUDIO macro, needed by dev/audio.c	Alexandre Ratchov
	ok miod and kettenis
2013-05-16	Run audio interrupts without grabbing the kernel lock. While IPL_SERIAL is	Mark Kettenis
	higher than IPL_AUDIO on sparc64, and interrupt handlers running at IPL_SERIAL still grab the kernel lock, this isn't a problem as machines with audioce(4) are guaranteed not to have zs(4) (the only driver that uses IPL_SERIAL).