summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2005-08-05syncTheo de Raadt
2005-08-05expose the algorithm used for splitting names, so that ArcCheck will beMarc Espie
able to use it.
2005-08-05minor KNF;Xavier Santolaria
2005-08-05activate sanity check: only files have !0 size.Marc Espie
2005-08-05fix splitting bugMarc Espie
2005-08-05Make sure $totsize != 0, even if the package does not contain any actualMarc Espie
files, so that the progress meter will work.
2005-08-05Make those tests match the actual implementation, not a not yet commited futureHans-Joerg Hoexer
version.
2005-08-05Correct URL. Noticed by Dries Schellekens <gwyllion at ulyssis.org>.Jonathan Gray
2005-08-05Some basic regression tests for ipsecctl, not connected yet.Hans-Joerg Hoexer
2005-08-05make three functions non-static (namespace is no issue, they might getDaniel Hartmeier
inlined), ok art@
2005-08-05tweak CAVEATS;Jason McIntyre
ok jsg@
2005-08-05impelmentation -> implementationMarc Espie
yes, new typo for an old friend.
2005-08-05strip multiple leading slashes from absolute paths; ok beck@Damien Miller
2005-08-05X40 has to be prodded a bit more to work.Jonathan Gray
tested by jolan@
2005-08-05List aps(4) in device lists. Reminded by deraadt@Jonathan Gray
2005-08-05Add support for SCSI pass-through commands for enclosure and processor devicesMarco Peereboom
to expose ses(4) and safte(4) information. ok beck@, dlg@ put your ami pt diff in, deraadt@
2005-08-05Uncomment aps to help gather more testing feedback.Jonathan Gray
I'd like to hear from anyone running with this.
2005-08-05Add commented entry for aps (ThinkPad Active Protection System)Jonathan Gray
2005-08-05Start of a driver for the "Active Protection System" accelerometerJonathan Gray
found in recentish ThinkPads. Sensors are reported via hw.sensors in sysctl but sensor numbers and presence is likely to change. Units and use of all of the sensors is not yet known as IBM has not released any documentation. Driver is based on information reverse engineered from the windows driver by Anurag Sharma and Mark Smith. "commit the code" deraadt@
2005-08-05typoTheo de Raadt
2005-08-05report the fan speed as a percentage since it seems all the enclosures weDavid Gwynne
fill in the rpm fields. your call deraadt@
2005-08-05add a sensor type that can report percentageDavid Gwynne
ok deraadt@
2005-08-05sync usage to man pageTheo de Raadt
2005-08-05flash up the man pageTheo de Raadt
2005-08-05if no command specified, assume -iTheo de Raadt
2005-08-05dont do an inquiry in attach just to print the safte revision number. itDavid Gwynne
is window dressing. zap it deraadt@
2005-08-05switch safte from using a scsi cmd at softclock to doing the commandsDavid Gwynne
in a kernel thread. very similair to the same change in ses.c. ok marco@ beck@ deraadt@
2005-08-05reorder the allocation of bits in attach to make cleaning up on an errorDavid Gwynne
shorter and less prone to mistakes ok marco@
2005-08-05READ BUFFER is a generic opcode, so dont define our ownDavid Gwynne
2005-08-05a define for the scsi READ BUFFER opcodeDavid Gwynne
for and ok marco@
2005-08-05shorten dmesg outputTheo de Raadt
2005-08-05scsi_safte.c and scsi_safte.h are now known as safte.c and safte.hDavid Gwynne
respectively ok krw@ deraadt@ marco@
2005-08-05make the naming of the safte files consisten with the rest of the scsiDavid Gwynne
drivers ok krw@ deraadt@ marco@
2005-08-04allow dcphy_status() to be called even when the parent interfaceBrad Smith
is in the down state. on some dc's with dcphy this will now allow the card to show carrier state and some link speed though not necessarily full-duplex for some reason. something to look into still.
2005-08-04syncTheo de Raadt
2005-08-04instead of static locals, malloc/free. the goal is to reduce stack usage,Daniel Hartmeier
but statics are dangerous in case of concurrency. ok deraadt@
2005-08-04activate libstubs (setlocale + dbopen)Marc Espie
okay deraadt@
2005-08-04zap rcsid. Okay deraadt@, krw@Marc Espie
2005-08-04fix worst three kernel stack pigs in xfs by moving stack allocation ofBob Beck
xfs message structures to malloc M_TEMP. ok art@
2005-08-04no cac(4) here.Brad Smith
2005-08-04this file is now known as ses.hDavid Gwynne
ok krw@ marco@ beck@
2005-08-04HW_PHYSMEM and HW_USERMEM should be unsigned; ok tdeval marcoTheo de Raadt
2005-08-04cac(4) are 2*, 3* and 4* series controllers.Brad Smith
2005-08-04- Compaq call it "Smart Array"Jason McIntyre
- 5/6 support, not just 5 reported by alexey e. suslikov; ok mickey@
2005-08-04the generic scsi attach line already says we're an enclosure, so why didDavid Gwynne
i make it say the second line that says the same thing?
2005-08-04be nicer to the controller the ses device is attached to by doing the scsiDavid Gwynne
commands in a kernel thread, not out of softclock with a timeout. help from pascoe@ ok marco@
2005-08-04better description of where in the sysctl tree the sensors are kept, andDavid Gwynne
a description of how ses matches tweaks from jmc@
2005-08-04buf_* was taken from the framework I wrote for bgpd (also used by ntpd, ospfdHenning Brauer
and some more). Here, however, we have a blocking socket for the IPC. buf_close (writes out) and buf_read can be simplified due to that, and they can just error out on short writes/reads. ok krw moritz
2005-08-04better wording, from jfb@. Ok jmc@.Xavier Santolaria
2005-08-04add and document import -d option; ok jmc@ joris@.Xavier Santolaria