summaryrefslogtreecommitdiff
path: root/sys/dev
AgeCommit message (Collapse)Author
2011-03-06Make sure MEMORYBARRIER(SYNC_REG) performs a read/write bus_space_barrier,Miod Vallat
instead of a read barrier only, as it is used in read-modify-write cycles. ok mjacob
2011-03-05Wait until the DMA engine is stopped before unmapping buffers and descriptors.Mark Kettenis
Fixes DMA errors seen on sparc64. ok miod@
2011-03-04Add TouchScreen support. Makes newer Gunze USB TouchScreen devices work.Mark Kettenis
ok deraadt@
2011-03-04Apparently keeping some USB devices in reset for too long makes themMark Kettenis
non-functional. Reduce the root port reset delay to 100 ms. This is still twice as long as the delay mandated by the standard, but makes the Gunze TouchScreen on the Sony Vaio VGN-UX91S and a few other devices work. ok jakemsr@
2011-03-04gather most pci conf register manipulation into a new function,Jacob Meuser
azalia_configure_pci(). use that function in azalia_attach() and azalia_resume(). fixes post-resume "static" during playback for some machines with ATI chipsets.
2011-03-04Peek at the interrupt status register before poking with the reply postMike Belopuhov
queue. In some situations this prevents us from reading a garbled reply. If this commit breaks your mpi, please report ASAP. The issue was reported and the fix was verified by Emeric Boit. Thanks! Ok dlg, kettenis, marco (who warned us and wanted to test more)
2011-03-03Prevent division-by-zero when scaling down to the screen resolution. MakeMark Kettenis
the WSMOUSEIO_SCALIBCOORDS fail with EINVAL if the settings would cause a division-by-zero. ok deraadt@
2011-03-02- change recieved -> received in comments, as well as two unused definesJasper Lievisse Adriaanse
in pxe.h. ok deraadt@ sthen@
2011-03-01back out r1.162, the one that bumps openings up on sas and fcDavid Gwynne
devices. my theory is that some devices report queue full conditions in ways the firmware doesnt understand, or some firmwares default to NOT doing the queue full handling internally. either way it reports queue full conditions as faulted io which gets passed up to the block layer as errors. this makes us conservative again and safe. this fixes panics from ajacout ok sthen@ deraadt@
2011-02-24we want to look up the device attached to a raid volume by its address onDavid Gwynne
the bus, not which bioctl volume id it is. ok mikeb@ krw@ miod@
2011-02-22Mark descriptors as invalid *before* unmapping the associated mbuf.Mark Kettenis
Might prevent the unrecoverable DMA error mentioned in PR 6557. ok damien@
2011-02-22Disable PCIe Active State Power Management (ASPM) to avoid bge(4) randomlyRobert Nagy
detaching from the bus on some of the machines. tested by many, ok miod@
2011-02-22In drm_modeset_ctl make sure to check the (now signed) value is > 0 asOwain Ainsworth
well as < the array length. ok kettenis@, mikeb@ (who suggested the same fix) and marco@. "get that fixed" deraadt@.
2011-02-22ami figures out the names of the enclosure attached to it by snoopingDavid Gwynne
commands on the passthru busses after the kernel becomes !cold. however, if the enclosure doesnt have any sensors then we dont issue regular commands to it, so ami cant see what its name is. this moves the name fetch from the passthru io path to a walk of all the devices on the bus just after they were attached. problem reported and fix tested by henning@ ok miod@
2011-02-21Make mos(4) set IFF_BROADCAST in if_flags.Stefan Sperling
ok jsg kevlo claudio; committing with miod's permission (src is soft-locked)
2011-02-21usage: mt [-f device] command [count]David Gwynne
most of the work and bug fixes were done by mikeb@, i put it in ioctl form. ok mikeb@ and miod@ tested by me an earlier revision was tested by Łukasz Czarniecki who also reported the problem in the first place.
2011-02-18alc_newbuf is always called from the interrupt context so it can't sleep;Mike Belopuhov
tested by Gabriel Linder, ok kevlo, miod
2011-02-17if changing the default speaker connection, be sure it's both aJacob Meuser
valid connection and a valid converter. fixes speaker auto-mute for some IDT codecs. ok miod@
2011-02-17allow interfaces other than slaves of the union interface to be theJacob Meuser
data interface. lets an "MTK GPS Receiver" work, as reported by Daniel C. Sinclair. ok miod@
2011-02-17the number of widget connections isn't necessarily the same as theJacob Meuser
number of widget connection list entries. if the high bit is set in a connection list entry, all widgets between the previous and current entry are connected. go through the connection list entry once to figure out how many connections there are, then go through it again to build the connection list. ok miod@
2011-02-15Initialize sc_iot early enough to avoid a crash when the hardware monitorMiod Vallat
function was found disabled. From NetBSD
2011-02-15A vnd buf doesn't only go through the vnd bufq but also the bufqThordur I. Bjornsson
for the device on which the vnd image resides on, this has the effect that a bufq_done call is done in the context of the underlaying bufq, setting b_bq to NULL, meaning there is never a bufq_done call done for the vnd bufq so the outstanding count never decreses. Add one in vndiodone. This fixes the suspsend issues krw@ was running into. ok tedu@, krw@ nod from miod@ on the commit.
2011-02-15Add workaround for BCM5906 A0/1/2 controller silicon bug. WhenRobert Nagy
auto-negotiation results in half-duplex operation, excess collision on the ethernet link may cause internal chip delays that may result in subsequent valid frames being dropped due to insufficient receive buffer resources. The workaround is to choose de-pipeline method as a flow control decision for SDI. De-pipeline method allows only 1 data in TxMbuf at a time such that a request to RDMA from SDI is made only when TxMbuf is empty. From FreeBSD; ok miod@; ok kettenis@
2011-02-15Do not compile statistics counters code if defined(SMALL_KERNEL). Makes i386Miod Vallat
RAMDISKA breathe a bit better. (file forgotten in yesterday's if_em.c commit)
2011-02-15bus_dmamap_sync() freshly initialized Rx descriptors before flipping the bitMark Kettenis
that hands them over to the hardware. This prevents the hardware from seeing stale contents if the compiler decides to re-order stores or if the hardware does store-reordering.
2011-02-15Copy pasto in comment. Fix by brad@ OK dlg@Claudio Jeker
2011-02-13Do not compile statistics counters code if defined(SMALL_KERNEL). Makes i386Miod Vallat
RAMDISKA breathe a bit better.
2011-02-11Port over a fix from Chris Wilson to poke the 855GM GWB/HIC bit.Marco Pfatschbacher
(https://bugs.freedesktop.org//attachment.cgi?id=41531) Fixes the heavy rendering errors and crashes that made the intel driver unusable on my 865G since 4.8 (kernel/6517) OK matthieu, oga.
2011-02-10now that usb_wait_task() is back, bring back recently revertedJacob Meuser
changes: * use usb_ref_{incr,decr,wait}() to avoid detaching the driver while a process is still using the hardware. * don't add timeout(9)s if the device is detached. * add checks to see if the device has been detached before running ioctls, timeouts, and tasks. * use usb_wait_task() to wait for tasks to complete. ok damien@
2011-02-09bring back changes in usb.c revision 1.73Jacob Meuser
date: 2011/02/04 08:21:39; author: jakemsr; state: Exp; lines: +85 -34 * merge 'onqueue' and 'running' members of struct usb_task into a new member, 'state'. * add new function 'usb_wait_task()', which waits for queued or running usb_tasks to complete. * in the USB_DEVICEINFO ioctl, fill struct usb_device_info in a usb_task, thereby avoiding races against driver attach/detach. but this time make the state a bitmask, since it is valid to be added to the task queue while the task is running. also be more careful about waking up the task when state changes. ok miod@
2011-02-09revert usb.c to r1.72, and all subsequent changes that depend on it.Jacob Meuser
this is causing problems with suspend/resume for some people.
2011-02-06cap the max size of the eeprom to 16k, as per the linux and freebsdDavid Gwynne
drivers. fix found by jonathan matthew and tested on an "Intel PRO/1000 (82576)" rev 0x01 which didnt work before this change. tested by jsg on the following chips: "Intel PRO/1000MT (82540EP)" rev 0x03 "Intel PRO/1000MT (82541GI)" rev 0x05 "Intel PRO/1000 MT (82574L)" rev 0x00 "Intel EP80579 LAN" rev 0x01 "Intel PRO/1000 PT (82572EI)" rev 0x06 "Intel PRO/1000 (82576)" rev 0x01 "Intel 82578DM" rev 0x06 "Intel I340-T4 (82580)" rev 0x01 some style issues are still being discussed which can be fixed in tree if necessary. ok jsg@
2011-02-05* use usb_ref_{incr,decr,wait}() to avoid detaching the driver while aJacob Meuser
process is still using the hardware. * don't add timeout(9)s if the device is detached. * add checks to see if the device has been detached before running ioctls, timeouts, and tasks. * use usb_wait_task() to wait for tasks to complete. ok damien@
2011-02-04* merge 'onqueue' and 'running' members of struct usb_task intoJacob Meuser
a new member, 'state'. * add new function 'usb_wait_task()', which waits for queued or running usb_tasks to complete. * in the USB_DEVICEINFO ioctl, fill struct usb_device_info in a usb_task, thereby avoiding races against driver attach/detach. ok miod@
2011-02-03Initialize stack memory before use. Found with clang.Matthew Dempsky
ok dlg@
2011-02-01Add the Toshiba 3G HSDPA MiniCard, found in my Dell Latitude D630.David Coppa
Ok jsg@, sthen@, miod@
2011-02-01regenDavid Coppa
2011-02-01Add the Toshiba 3G HSDPA MiniCard, found in my Dell Latitude D630.David Coppa
Ok jsg@, sthen@, miod@
2011-01-30another runTed Unangst
2011-01-30regenTed Unangst
2011-01-30another run device. ok miodTed Unangst
2011-01-30Report the compose key on sun keyboards in raw mode as a different keycodeMiod Vallat
from right alt. ok krw@ shadchin@ deraadt@
2011-01-29spaces that make my eyes bleed. no binary change.Marco Peereboom
2011-01-29Fix two logic errors:Kevin Lo
- "could not disable Rx/Tx MAC" from FreeBSD - "could not disable RxQ/TxQ" from Gabriel Linder
2011-01-28we deal with polled ccbs erroring out properly, we dont have to complainDavid Gwynne
about it unless we want verbose debug shizz.
2011-01-28ahci0: timed out 1, active 0, count 0 during pmp probe timeouts should beDavid Gwynne
quiet unless we want verbose output.
2011-01-27Fix newlines for error on attach case. with & ok dlg.Darren Tucker
2011-01-27while closing the device, don't sleep with the PCATCH flag, otherwiseAlexandre Ratchov
a signal (like SIGALRM) may cause the output buffer to not be drained and causing for instance stale notes.
2011-01-26Add port multiplier support, has been in snaps for a while with no reportedDale Rahn
issues. No actual OKs, but general acknowledgement and 'get it in' from several.
2011-01-26syncKevin Lo
generated by cgit v1.2.3 (git 2.25.1) at 2025-01-19 11:28:20 +0000