summaryrefslogtreecommitdiff
path: root/sys/dev
AgeCommit message (Collapse)Author
2008-11-04The drawable spinlock is no longer ever held in interrupt context (andOwain Ainsworth
won't be). It doesn't need to block interrupts anymore so switch to IPL_NONE.
2008-11-04Enable IMR passthrough of vblank events before enabling it in theOwain Ainsworth
pipestat register. Fixes a nasty race where the bit would get set without being reflected in the interrupt register, so we'd never get another vblank interrupt. Also, use the user_irq_lock to also protect vblank register writes, since it covers the same register. From Eric Anholt and Keith Packard at Intel.
2008-11-04Hold the drm lock around all things that touch the ringbuffer.Owain Ainsworth
Not strictly needed in the non-gem case, but it will be needed then, and doesn't hurt now. From Eric Anholt at intel.
2008-11-04revert the pageflipping and vblank sync code to the older style thatOwain Ainsworth
doesn't handle triple buffering (which has been marked ``don't use this, it's unstable'' for ever anyway) While the code just removed is in drm git. it's not planned to go any further, due to being a horribly ugly hack. Instead a proper fix which will depend on memory management is planned. So revert this stuff here, since it's now dead. Testing shows no regressions.
2008-11-03HW CCMP encryption/decryption of unicast frames.Damien Bergamini
HW support for other ciphers (TKIP) really sucks (no Michael MIC, not even TKIP Phase 1) and is thus not implemented here. fix a panic in iwn_fatal_intr() when dumping the firmware error log while i'm here.
2008-11-03regenRobert Nagy
2008-11-03add Radeon HD 3650Robert Nagy
2008-11-03Don't attach bio unless we do RAID.Marco Peereboom
2008-11-03support w83627dhg at i2c (as seen on Sun Ultra 24 via dmesg@) ok deraadt@Constantine A. Murenin
2008-11-02rcs tagsJonathan Gray
2008-11-02Remove second commented out call to ether_ifdetach() inJonathan Gray
mos_detach().
2008-11-02Remove duplicate call to ether_ifdetach() in axe_detach()Jonathan Gray
which has been there since rev 1.1. Spotted by Johann Christian Rode <jcrode@gmx.net>
2008-11-02Start updating isp. Just new defines, comments, whitespace, anything thatKenneth R Westerback
doesn't change the .o. Use isp_stds.h.
2008-11-02Start updating isp. Just new defines, comments, whitespace, anything thatKenneth R Westerback
doesn't change the .o.
2008-11-02Start updating isp. Just new defines, comments, whitespace, anything thatKenneth R Westerback
doesn't change the .o.
2008-11-01Add sensorsMarco Peereboom
2008-11-01Fix bogus shift.Marco Peereboom
2008-11-01Start updating isp. Just new defines, comments, whitespace, anything thatKenneth R Westerback
doesn't change the .o.
2008-11-01Oops forgot .h part.Marco Peereboom
2008-11-01Add disk. This makes bio mostly done however to make it pretty we needMarco Peereboom
to implement RAID_ACTION. Remains disabled for now.
2008-11-01new workq_create interface. discovered by Vladimir KirillovTed Unangst
2008-11-01set the appropriate value (0, not emulated) in the flags field ofJacob Meuser
struct audio_encoding in azalia_query_encodings. otherwise it will be some random value. makes AUDIO_GETENC and thus sio_getcap(3) much more useful here.
2008-10-31clean up error messages as suggested by jakemsrTheo de Raadt
2008-10-31Generalize function to obtain all details to do bio stuff. This is inMarco Peereboom
preparation to start showing unsused disks which will require all this information to do the math. While at it fix a size of hotspare volume bug.
2008-10-31Add opcodeMarco Peereboom
2008-10-31Remove the vblank_pipe member of the softc and noop the ioctl that setsOwain Ainsworth
it. Since the vblank handling got reworked this is very much unneeded.
2008-10-31Remove another interface (this one for ARB_Occlusion_Query) that we'veOwain Ainsworth
never used (mesa did a version check that never passed on our kernel). The correct way to do this is in mesa master and requires memory management.
2008-10-31Handle "jack sense" hp/speaker switching by polling for an interruptJacob Meuser
instead of using unsolicited events. Already supported codecs with custom unsolicited event handlers are not affected, yet. from Alexey Suslikov, thanks
2008-10-31Support for converter groups where members support non-identical sampleJacob Meuser
sizes and rates. from NetBSD via brad@ Also, * use a single variable to hold the number of format variations (from NetBSD) * initialize a variable (f->frequency_type) before use (from NetBSD) * fix a typo from Alexey Suslikov, thanks
2008-10-31some leftover bits from r1.32 that should have been removed inJacob Meuser
r1.51 pointed out by Alexey Suslikov, thanks
2008-10-30regenBrad Smith
2008-10-30Add the PCI id for the Promise PDC42819 chipset.Brad Smith
2008-10-30Bunch of cleanup. Kill some dead variables and some stupid code.Owain Ainsworth
Some ideas taken from upstream.
2008-10-30Fix bug that when there are no volumes but there are hot spares it wouldMarco Peereboom
no show the hotspares.
2008-10-30From Jeremy Morse via NetBSD:Marcus Glocker
* Serialize access to the ehci intrlist. * Change the ehci intrlist to a tailq so xfers are not queued out of order. * In ehci_check_itd_intr, don't treat a transfer error as an indication that the xfer is no longer active. This also fixes "ehci_allocx not free" errors seen recently. Tested and OK brad@, kevlo@
2008-10-30for devices that do not allow independent setting of recording andJacob Meuser
playback parameters (like the macppc i2s devices), be sure to synchronise recording parameters to playback parameters when configuring for full-duplex operation, as audio(4) dictates. ok ratchov
2008-10-29add a flag to know is this is Audigy 2 ValueJacob Meuser
2008-10-29support for Audigy 2 Value based on patch from Dennis SuhoninJacob Meuser
(openlunatic at gmail). tested by Dawe (dawedawe at gmx de), thanks, and me. does not disturb currenlty working emu(4) devices.
2008-10-29Add initial bio support; only volume status is handled.Brad Smith
Based on cac(4) bio support from NetBSD which is derived from OpenBSD's ciss(4) bio support. Tested by otto@ jbg@ ok marco@
2008-10-29Bring pd structure up to dateMarco Peereboom
2008-10-29- Add a workaround field to the softc struct.Brad Smith
- Move the extended FIFO workaround from the capabilities field to the workaround field. - Add General purpose register 1 defines for the workarounds. - Enable a workaround for CRC errors that can be experienced with A2 revision adapters. - Add a workaround for packet loss that can be experienced with A2 revision adapters when in 10/100 mode with half duplex. From DraonFly - Rename the workaround flags so they describe what is being worked around. ok jsg@
2008-10-29dlg says "well, that is embarassing"Theo de Raadt
2008-10-28make DRM_INFO() a bit less gross, though getting all the grossness out at ↵Theo de Raadt
once is impossible; ok oga
2008-10-28#if NVLAN > 0 in one more spot.Brad Smith
2008-10-28Fix output when reporting on multivolume disk groups.Marco Peereboom
Reported by mitja@muzenic.net
2008-10-28Add the MacBookPro4,1 to the list of systems with GPIO quirks andBrad Smith
needs adjustment. Allows the integrated speakers to work. from Alexey Suslikov
2008-10-28#include sensorsMarco Peereboom
2008-10-28Shut splassert upMarco Peereboom
Fix an undo error in previous commit
2008-10-28Be nicer waking up.Marco Peereboom
prompted and ok dlg
2008-10-28Major overhaul of bio.Marco Peereboom
Fix set hotspare that didn't always work. Fix several very subtle bugs because of firmware lies Fix disk size that sometimes was incorrect Only poke drives if something changed so it makes bio way faster on subsequent calls Tested by several folks, thanks Ok dlg
generated by cgit v1.2.3 (git 2.25.1) at 2025-01-29 13:59:44 +0000