| Age | Commit message (Collapse) | Author |
|---|
|
MSDOSFS code. Eliminates -G option to mount_msdos.
Nit detection by gwk@, tom@, jmc@.
ok weingart@ tom@ thib@ dlg@ deraadt@
|
|
on old world macs as the external SCSI controller. Still a work in
progress, it probes the SCSI bus correctly but attempts to access device
meet with limited sucess.
|
|
number, it can differ from the sequence number being tested (for packets
without payload), and both matter in explaining why a packet mismatched.
|
|
domain number such that we can distinguish between busses with the same bus
number that are behind different host bridges. Domains can be accessed by
using different device nodes.
ok deraadt@
|
|
routing information.
|
|
|
|
process that re-uses its ASID from getting random SIGTRAPs.
ok miod@, mickey@
|
|
ok weingart@ pedro@
|
|
|
|
type. this would cause us some more trouble with WME which isn't
supported yet...
thanks to Nick Kossifidis
|
|
represented in the FAT, limit the number of clusters we work with
to the FAT value. This stops corrupt filesystems causing us to run
off the end of the FAT and panic()ing in fillinusemap().
Found by Jason Crawford (jasonrcrawford at gmail.com) with the MOKB
fs fuzzer. Initial debugging by thib@.
ok krw@
|
|
likely that interrupts are properly counted in kern.cp_time sysctl. so top(1)
et al will show a value for time spent in interrupts.
tested by many, ok drahn@
|
|
register offsets, loose the file meshreg.h.
ok dlg@
|
|
adding additional IP address or re-running dhclient by only causing an
interface init if the interface is not already running. As well as some
cosmetics.
from brad@.
Loose the ether_cmp function of dubious optimizational value simply
use bcmp.
Revert the ordering hack, don't execute instruction eieio directly in
mc_dmaintr, makes no difference in triggering the uvm_fault bug, the
problem lies elsewhere.
Fix some indentation related to the DBDMA macros.
Tested on Power Macintosh 9600MP and 9500MP by myself.
|
|
doing nothing and the mi code does not use it anymore anyway. No functional
change.
|
|
waste of 128KB of kvm.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
diff from Sepherosa Ziehau (DragonFly); The firmware TIM element template
was initialized wrong.
Though, our ieee80211_alloc_beacon() routine calculates it's own TIM
element, which now results in having two TIMs in an acx(4) beacon.
Would it be an idea to introduce a ieee80211_alloc_beacon() flag to
turn the TIM element off? Because fiddling out the TIM element in the
driver for cases in which the firmware calculates its own TIM is
a bit ugly ...
ok claudio@
|
|
|
|
(s6_addr16[1] filled)
ok dhartmei
|
|
reuses IPv4 signature file (assuming that TCP code is shared among IPv4/v6).
mcbride ok.
|
|
errors to userland: make all cpu_setperf functions return void.
Tested by many, ok gwk@
|
|
|
|
in a much more recent Marvell Linux sk98lin driver.
From brad@.
ok me and kettenis (via mglocker)
|
|
|
|
|
|
|
|
|
|
While there, fix a typo.
ok gwk
|
|
- change the defs of LHS, RHS, DST and DST2 to be AML_*.
Noted by Marco S Hyman on misc@, these allow ACPI to be compiled into
RAMDISK kernels. Suggestion for AML_ rename (to avoid name clash with
TIMEZONE's DST, which was noted by someone else a while ago, but I
forget who) from jordan@.
ok marco@, jordan@
|
|
ok otto@
|
|
hardware checksum could be mangled with 28 byte or less IP packets.
>From tsutsui@NetBSD
Tested by otto@ brad@ and Han Boetes <han at mijncomputer dot nl>
Tested by maja@ with the 8139C+ chipset
Tested on amd64/i386/sparc64
ok mglocker@ brad@ otto@
|
|
-- bus_space(9) does the endian switch for us -- and finaly change one
letoh32() to letoh16(). The last one being by far the nastiest one.
OK mglocker@
|
|
|
|
using the new NO_CCB result. Currently a no-op since no driver produces
that result.
ok dlg@ marco@ deraadt@
|
|
drop it. im guessing at some point we'll want to use each port, knowing
where they are can help.
|
|
|
|
ahci_port structs for it.
|
|
ccb lists. this is missing handling of the dmamem each port needs to
manage, but that can wait till i figure out how it fits together. insane
stuff.
|
|
|
|
|
|
and ahci_pwrite. these will use an io handle into a window of the
controllers register space that can be created with bus_space_subregion,
and which is stored in the ahci_port struct.
|
|
per port rather than as a big list for the whole controller. the softc
has an array of pointers to a struct that manages all the shizz relevant to
a particular port. when we map the controllers ports, we'll allocate these
as needed.
if the port isnt implemented, we leave its pointer in the softc null and
use that to decide if a drive might be there or not, rather than lookup up
the PI register all the time. a comparison has to null has to be cheaper
than doing bit operations, surely.
|
|
|
|
introduced by myself at h2k6, thanks mickey@ and toby@ for turning me in
the direction of a corrupt PCB.
|
