summaryrefslogtreecommitdiff
path: root/sys/arch
AgeCommit message (Collapse)Author
2004-02-02PSL_XCRYPT is 0x40000000Theo de Raadt
2004-02-02I think there is no reason to splhigh around the VIA crypto stuffTheo de Raadt
anymore; also, indicate to userland that it may now safely use the instructions
2004-02-01More work in attempt to get interrupts working reliably:Miod Vallat
- do not update 8259 masks after every modification, but let the code responsible for the changes update it at the most appropriate time. - be sure to correctly acknowledge interrupts on the 8259. - and various cleaning. This still fails when interrupts come too fast, at the moment.
2004-02-01set i386_has_xcrypt at the right place, which exposes itself asTheo de Raadt
sysctl machdep.xcrypt Do not enabled yet..
2004-02-01Use "pushfl; popfl" sequence before each xcrypt-* instruction. AccordingTheo de Raadt
to the manual, any load into the EFLAGS register clears bit 30, resulting in key reload. This is the mechanism that permits multi-process use of the xcrypt-* instruction..
2004-02-01use VIA xstore-rng and xcrypt-* instructions, now that gas groks themTheo de Raadt
2004-02-01via crypto code no longer needs to temporarily enable FSXR -- it isTheo de Raadt
now always on.
2004-02-01Merge many freebsd diffsTheo de Raadt
add VIA c3 crypto instructions add fxsave, fxrstor
2004-02-01Enable FXSR all the time, and cope with NPX/FXSR conversions; from netbsd.Theo de Raadt
Pass SSE/SSE2/XCRYPT flags out via syctl, and prepare for being able to do xcrypt-* in userland; ok naddy, tested a lot by pvalchev and jolan, also works on amd64 in 32bit mode
2004-02-01NetBSD -> OpenBSD in the comments.Miod Vallat
2004-02-01Trivial touches to compile with gcc 3.Miod Vallat
2004-02-01Do not compile ptrace-related functions unless option PTRACE.Miod Vallat
2004-02-01Do not compile ptrace-related functions unless option PTRACE.Miod Vallat
2004-02-01Sync maxpartitions value with <machine/disklabel.h>Miod Vallat
ok drahn@
2004-02-01Sync user ldt code with NetBSD:Alexander Yurchenko
- finally remove it from pcb, it's a pmap thing only - more sanity checks - better lockin - may be something else Fixes panics when using apps requiring it (mplayer-win32 e.g.). Problem found and test espie@. OKs from miod@ (sshhh, don't tell anyone) and art@.
2004-02-01tag...Dale Rahn
2004-02-01fix tags...Dale Rahn
2004-02-01cats (ARM) support for OpenBSD, based on NetBSD code.Dale Rahn
2004-02-01Arm port, NetBSD codebase stripped down, 32bit only support.Dale Rahn
2004-01-31add another USB device... as requestedTheo de Raadt
2004-01-31rename SIMD/SIMD2 to SSE/SSE2Theo de Raadt
2004-01-30Remove redundant initialization.Miod Vallat
2004-01-30Some cleaning and sync to the macppc openpic code.Miod Vallat
No functional result.
2004-01-29typos; jjy2+@pitt.eduTheo de Raadt
2004-01-29typos; jjy2+@pitt.eduTheo de Raadt
2004-01-29Pretty printf.Miod Vallat
2004-01-29Have the bootloader pass RB_ flags, as well as the boot device, to the kernel,Miod Vallat
and have the kernel process them correctly. Right now code to match the BUG's idea of the boot device with the kernel's is missing.
2004-01-29Print system type early in autoconf, as determined from the board statusMiod Vallat
register.
2004-01-29Some more cleaning and sync with other ppc ports, in order to help with aMiod Vallat
future factorization of the pci bridge code.
2004-01-29In bus_dmamap_sync(), be sure to flush a host pa.Miod Vallat
2004-01-29as seen in freebsd: asm pagezero implementations, but use a fn pointer.Ted Unangst
the sse2 version cuts ZOD fault time in half. suggestions mickey deraadt. many many testers. ok deraadt.
2004-01-29check struct reg for evil; from netbsdMichael Shalayeff
2004-01-29struct regs for realMichael Shalayeff
2004-01-29Don't assume 1:1 mapping for bus_spac anymore.Miod Vallat
2004-01-29Map the raven, pcic, and isa/pci io space virtual rather than 1:1.Miod Vallat
2004-01-29Allow "halt" as a root device, to return to the BUG, as done on some otherMiod Vallat
platforms.
2004-01-29Simplify test for CPUID_TSC in cpu_featureTom Cosgrove
ok fgsch@, deraadt@
2004-01-29No longer accept memory ranges in KB from /boot, which hasn't doneTom Cosgrove
this since 1999. If /boot tries to, we will now panic() with the "/boot too old" message. No-one should be using such an old /boot anyway, since it didn't support ELF kernels then. ok weingart@, deraadt@
2004-01-29Correctly handle 88110 exceptions occuring in a delay slot: control mustMiod Vallat
be restored with valid EXIP+D and ENIP values in this case, unless it is decided to skip instructions.
2004-01-29187BUG, 188BUG and 197BUG agree that only SSR3 needs to be preserved acrossMiod Vallat
BUG calls, though SSR0-SSR2 will be corrupted on some BUGs. This allows up to simplify slightly the context switching code.
2004-01-28Return to BUG if main() returns, rather than falling through the scheduler...Miod Vallat
2004-01-28Do not put the sample IndustryPack driver in GENERIC. This makes no real sense.Miod Vallat
2004-01-28Correctly protect against multiple inclusion...Miod Vallat
2004-01-28More clock handling cleanup:Miod Vallat
- move clock callbacks outside the firmware structure - map the clock virtual. A bit tricky because it needs to be mapped before autoconf... - compensate for the removal of 1:1 segment 8 mapping
2004-01-28Better fix the typo for real, this time. Oops.Miod Vallat
2004-01-28Use write-back mappings whenever possible again, but make sure thatMiod Vallat
vmapbuf() returns write-through mappings.
2004-01-28Add tftpboot to the mvme88k builds.Miod Vallat
2004-01-28Skip first argument on commandline, so we can try to actually boot theMiod Vallat
requested kernel rather than ourselves.
2004-01-28sniff. first file i add'd is first to be rm'dTed Unangst
2004-01-28duh. no need for swapgeneric here too; from deraadt@Michael Shalayeff
generated by cgit v1.2.3 (git 2.25.1) at 2025-01-26 18:19:11 +0000