summaryrefslogtreecommitdiff
path: root/sys
AgeCommit message (Expand)Author
2007-04-25more register init in tht_upDavid Gwynne
2007-04-25Yet another case that forgot to set the time.Artur Grabowski
2007-04-25Set the time even when the clock chip is bad.Artur Grabowski
2007-04-25put a two second timeout on the chip init after the firmware is loadedDavid Gwynne
2007-04-25shift some bits like the spec says i should.David Gwynne
2007-04-25format string fixDavid Gwynne
2007-04-25add some debug to the tx paths so i can see packets go on and off the hw.David Gwynne
2007-04-25fix the conditions that the tx free path keeps looping on so that it willDavid Gwynne
2007-04-25configure the 10G mac and enable (more) interrupts when the interface isDavid Gwynne
2007-04-25split tht_fifo_ready into tht_fifo_writable and tht_fifo_readable. the wayDavid Gwynne
2007-04-25start implementing interrupt handling.David Gwynne
2007-04-25add more register definitions, in particular the interrupt ones.David Gwynne
2007-04-25im an idiot. for (;;) { } while (condition); loops forever, no matter whatDavid Gwynne
2007-04-25add some debug goo to be used soonDavid Gwynne
2007-04-25Same change as i386, use the number of states to calculate performanceGordon Willem Klok
2007-04-24Pass a real dev_t to readdisklabel(), not a unit number with lipstick.Kenneth R Westerback
2007-04-24Tell init(8) to power down machine instead of just halting it.Mark Kettenis
2007-04-24Dont log() send failures in nfs_send(), they arent helpful since we retryThordur I. Bjornsson
2007-04-24do not printf+panic, when a longer panic message is easier; kettenis okTheo de Raadt
2007-04-24Make power(4) map registers in io space if mem space fails.Mark Kettenis
2007-04-24Choose the state for a request performance level based on calculationsGordon Willem Klok
2007-04-24It's never too late to write "too late" correctly.Miod Vallat
2007-04-24exeption -> exceptionMiod Vallat
2007-04-24Firmware lies! Switching from LEVEL to EDGE triggered interrupt preventsGordon Willem Klok
2007-04-24apic_address should be 32 bits, rather than paddr_t; prepares us forTom Cosgrove
2007-04-24Use the physical address we have stored for the page directory ratherTom Cosgrove
2007-04-24Fix mounting of ffs2 via "mount -a". ffs2 is not separate from ffsTodd C. Miller
2007-04-24do not use static function declarations in kernel drivers (noReyk Floeter
2007-04-24use the fifos ready byte counter in the firmware loading.David Gwynne
2007-04-23Add initial test for existing metadata on chunks.Marco Peereboom
2007-04-23Add interesting runtime stats.Marco Peereboom
2007-04-23When receiving an undock request, walk the list of dependant devicesMichael Knudsen
2007-04-23add ethernet type for "slow protocols" (defined in IEEE 802.3ad)Reyk Floeter
2007-04-23When adding a device node to the list of dependant devices, don't addMichael Knudsen
2007-04-23In acpidock_softc, change sc_sens from being a one element long array toMichael Knudsen
2007-04-23Keep track of coerced chunk sizes.Marco Peereboom
2007-04-23rename drivers with numbers in them; ok gwk drahnTheo de Raadt
2007-04-23Add an unbranded device I have, chip says its a PL-2303HX.Jonathan Gray
2007-04-23regenJonathan Gray
2007-04-23New unbranded uplcom(4).Jonathan Gray
2007-04-23iqg -> irqGordon Willem Klok
2007-04-23increment opackets and ipackets. clear OACTIVE when tx completes.David Gwynne
2007-04-23Clean up some comments.Artur Grabowski
2007-04-23spelling in comment fix.David Gwynne
2007-04-23when we use a fifo we always seem to need to know how much of the fifo isDavid Gwynne
2007-04-23Remove 'cg_space' from 'struct cg'. Due to the alignment on 64-bitPedro Martelletto
2007-04-23Switch alpha to timecounters.Artur Grabowski
2007-04-23replace the code in the rxf and txt fifo handlers that loads the dmaDavid Gwynne
2007-04-23add tht_write_dmap, and tht_write_pad.David Gwynne
2007-04-23Clean up an obsolete allocator.Artur Grabowski