summaryrefslogtreecommitdiffstats
path: root/sys/dev
AgeCommit message (Collapse)AuthorFilesLines
2014-10-08Remove the i830/i845 pipe A force quirks, matching changesjsg1-5/+1
in recent versions of the linux drm code. drm/i915: Nuke pipe A quirk on i830M From Daniel Vetter 0e8abc81bf4b54a01bb026ba4d883de7c415cc03 in mainline linux drm/i915: Undo the PIPEA quirk for i845 From Chris Wilson a4945f9522d27e1e6d64a02ad055e83768cb0896 in mainline linux With this change a hardware lockup with i845g reported by "THEvoid" via bugs@ is avoided.
2014-10-07remove preliminary AOE (ata over ethernet) support. not finished aftertedu3-912/+3
many years and wide spread demand for support never materialized. time to pack it in.
2014-10-07zap duplicated line; Theo Buehlersthen1-2/+1
2014-10-05Only synchronize used TRBs and not the full ring when sending a controlmpi1-9/+13
transfer. While here remove/fix other XXXs.
2014-10-05Do not mark the pipe as halted when the HC reports a (split) transactionmpi2-3/+7
error. Makes Intel Series 7 controllers happy and no longer report an illegal context state transition when detaching devices.
2014-10-05Do not re-establish the default pipe twice for every controller. Movempi3-28/+78
this hack in the drivers that need it. Tested by many, thanks! ok pirofti@, kettenis@
2014-10-05Quirk needed by the Archos 24y Vision, from Max Fillinger.mpi1-1/+9
2014-10-05regenmpi2-4/+16
2014-10-05Add vendor and product for the Archos 24y Vision, from Max Fillinger.mpi1-1/+5
2014-10-05two more events found on thinkpad yogaderaadt1-1/+6
2014-10-04- Recognize the Yukon Prime and Yukon Optima 2 chipsets.brad2-5/+52
- Recognize some more revs of the EC U and Supreme chipsets. - Add the PCI id for the Yukon 88E8079. ok jsg@
2014-10-04Wait until a read control transfer is really completed before passingmpi1-5/+24
it to the stack when a Short Transfer condition is reported. In this dummy implementation the ``Event Data TRB'' of a read control transfer is the only TRB that can trigger an interrupt without being the last TRB of a transfer. This is done in order to report the remaining length of a short transfer. But when that happens, we want to wait until all Transfer TRBs are completed before passing the xfer to the stack. Note that clearing the ISP and IOC flags in all Transfer TRBs like it is specified in 4.10.1.1.1 might not work in our cases because the HC has most of the time already processed all Transfer TRBs when the driver dequeues the events in the softinterrupt path. While here, use the right spl protection when aborting a xfer.
2014-10-04replace mutexes to serialise the operations on the flag that restrictsdlg1-17/+7
the number of contexts that are refilling the rx rings with atomic ops. this is borrowed from code i wrote for the scsi midlayer but cant put in yet because i havent got atomic.h up to scrach on all archs yet. the archs myx runs on do have enough atomic.h to be fine though.
2014-10-04syncderaadt2-2/+7
2014-10-04more nvidia; from Seth Jacksonderaadt1-1/+2
2014-10-03regenbrad2-2/+12
2014-10-03Add Atheros E2200 and Marvell Yukon 88E8079.brad1-1/+3
2014-10-03refill the rx ring in myx_rxeof, not much later at the end of myx_intr.dlg1-18/+7
2014-10-03in rxeof, instead of taking the biglock on every packet to call bpfdlg1-10/+17
and ether_input, queue all the mbufs onto an mbuf_list on the stack and then take the biglock once outside the loop.
2014-10-03we dont need the kernel lock to call bus_dmamap_load and unload thanksdlg1-7/+3
to ketenis. move the if_ipacket and if_opacket increments out of biglock too. theyre only updated from the interrupt handler, which is only run on a single cpu so there's no chance of the update racing. everywhere else only reads them.
2014-10-03dont need to hold the kernel lock to call MCLGETI and m_freem now.dlg1-6/+2
2014-10-03dont take the kernel lock on every interrupt in case we might changedlg1-4/+7
the link state or to clear OACTIVE, just take it when we know we really are going to do those things.
2014-10-01Check that the speed of a new device does not exceed its parent's speedmpi2-13/+9
prior to calling usbd_new_device(). This will let us set the specified default Max Packet Size before opening the control pipe and reduces spaghetti! ok stsp@
2014-09-30Fix another Ricoh SD controller which needs to be put into slow mode to work.stsp1-3/+5
reported by mlarkin@; ok mlarkin@ deraadt@
2014-09-30Add support for "physical devices". Tested on:kettenis1-2/+261
mfii0 at pci1 dev 0 function 0 "Symbios Logic MegaRAID SAS2208" rev 0x05: msi mfii0: "LSI MegaRAID ROMB", firmware 23.22.0-0012, 1024MB cache ok dlg@
2014-09-30prevent dmesg spam from some new windows-only keys found on very newphessler1-1/+10
thinkpads OK mpi@ miod@
2014-09-29correctly decrement the C_SCSIXFER reference count in wdc_xfer flagsdlg1-2/+5
if the ata stack xfer free path is called before teh scsi xfer free path. this fixes a leak. in the vast majority of uses (eg, filesystems mounted on cd(4) via atapiscsi(4)) the scsi xfer free is always done first. however, if you're doing ioctls (eg, cdio(1)), the scsi ref is held over to the ioctl completion path, so the ata xfer free happens first. problem found by patrick keshishian who did some excellent work debugging this and providing the right fix for it. ok jmatthew@
2014-09-27retire Accoom Networks Artery T1/E1 drive; ok claudioderaadt8-4317/+1
2014-09-27add some more sdtemp devicesjsg2-3/+63
ok deraadt@
2014-09-26Don't mask off revision bits for tests against a fulljsg1-3/+3
device id/revision id value. Makes it possible for se98 and stts424 sensors to match. ok deraadt@
2014-09-26new ricoh controllers must also be pushed into "old slow mode" at resumederaadt1-20/+59
time. ok kettenis jsg
2014-09-26Revert a part of r1.98 if_run.c which caused a regression on older genmlarkin1-45/+33
run(4) devices. ok stsp@
2014-09-26Fix format signednessguenther1-2/+2
ok blambert@ doug@
2014-09-26Rework piglet and pig allocation. Currently the piglet gets allocatedkettenis1-11/+14
deep down in the suspend path, where it is really hard to recover from allocation failure. So allocate the piglet early on in the suspend path. Also change the piglet and piglet allocation functions to use km_alloc(9) instead of doing pmemrange magic. This removes a bunch of code which, in the case of the piglet allocation, is broken since it results in a NULL pointer dereference. Also switch the piglet allocation to not wait. If we can't allocate 16MB of phys contig memory on a halfway modern machine we're almost certainly under a lot of memory pressure and we're better off not trying to hibernate anyway. ok mlarkin@
2014-09-25Bring back rev 1.249. Now that mp_setperf() has been fixed, it should bekettenis1-1/+7
safe to restore perflevel when resuming. Should fix the issues people see with the processor running at full speed heating up the machine while the apm command reports that the CPUs are running at a low speed. ok dcoppa@, deraadt@
2014-09-25include bpfilter.h so the softc will be the same size in all filesjsg2-2/+6
2014-09-24fix the build when DRMDEBUG is definedjsg1-2/+2
2014-09-24Kill two generic DVACT_DEACTIVATE handlers since this event is nowmpi2-18/+5
autoconf(9)ally passed to the children of a device. ok ratchov@
2014-09-24Replace a few malloc() calls with reallocarray() so it checks for overflow.doug1-2/+2
ok deraadt@
2014-09-23Enable MSI support.brad1-2/+2
Tested with a number of chipsets.
2014-09-23Release the acpi lock around calls to wsdisplay_suspend() andkettenis1-1/+9
wsdisplay_resume(). The X server may adjust the display brightness just before/after the VT switch when suspending/resuming. In fact, it may already be blocked adjusting the display brightness for some reason unrelated to the suspend request. The VT switch request will eventually time out, but the delay is annoying and the failed VT switch will lead to problems after resume. Needed for a future update of the xf86-video-intel driver. ok mlarkin@
2014-09-23add Bay Trail and Braswell SMBusjsg1-1/+3
2014-09-23regenjsg2-2/+12
2014-09-23add Bay Trail and Braswell SMBusjsg1-1/+3
2014-09-23Fix high capacity (> 2GB) eMMC support.rapha2-12/+18
Based on a diff by Cedric Tessier, nezetic at gmail dot com, thanks! Discussed with and ok jsg@
2014-09-23Make use of the red-black tree as the original Linux code did insteadjsg1-3/+6
of interating over the tree like a list. From Andriy Gapon in FreeBSD. ok kettenis@
2014-09-23dont need to call bpfdetach cos if_detach does it for you.dlg1-9/+1
delete some stupid comments while there. ok deraadt@ brad@
2014-09-22add Winchiphead CH382 support.sasano1-1/+17
CH382 has 2S/2S1P/1P configuration, the diff supports 2S/2S1P only (2S1P works as 2S). ok by deraadt@ and kettenis@
2014-09-22the kn_hook member of a knote is a void *, so we dont have to castdlg1-4/+4
to or from it. ok gcc
2014-09-21add a missing break statement in the RL_HWREV_8168E_VL casejsg1-1/+2
ok brad@
Copyright © 1996 – 2026 Jason A. Donenfeld. All Rights Reverse Engineered.