| Age | Commit message (Collapse) | Author | |
|---|---|---|---|
| 2009-12-14 | Fix previous commit. I left out a small fragment and accidentally committed | Mark Kettenis | |
| a change we're not readdy for yet. | |||
| 2009-12-14 | Support for in-band descriptor mode (VIO_DESC_MODE) used by OBP. | Mark Kettenis | |
| 2009-12-14 | Make sure interrupts are disabled when we don't have queues configured. This | Mark Kettenis | |
| prevents interrupt storms I'm hitting when running OpenBSD as a control domain. | |||
| 2009-12-14 | Add wrapper functions to enable/disable interrupts and manipulate their state. | Mark Kettenis | |
| 2009-12-13 | bit less aggressive about hiding; ok miod | Theo de Raadt | |
| 2009-12-12 | Protect more definitions userland doesn't need to care about with _KERNEL | Miod Vallat | |
| 2009-12-12 | Switch page size from 4KB to 16KB on R10k kernels without R5k support | Miod Vallat | |
| (i.e. IP27 and IP30 sgi kernels). | |||
| 2009-12-12 | Introduce various CPU_### options to enable cpu-specific code or workarounds; | Miod Vallat | |
| be sure to rerun config(8) on your kernel configuration files. | |||
| 2009-12-12 | unifdef -D__LP64__ | Miod Vallat | |
| 2009-12-12 | Add missing splbio() protection when sending commands to the vDisk server. | Mark Kettenis | |
| 2009-12-12 | Remove debug printf and properly dequeue command instead when a read times out | Mark Kettenis | |
| on LOMLite2. | |||
| 2009-12-11 | By popular demand and peer pressure, check-in work in progress work to support | Miod Vallat | |
| the Yeelong Lemote mips-based netbook. Kernel bits only for now, needs polishing; most of this work done during h2k9 last month, although the porting effort started earlier this year. | |||
| 2009-12-11 | By popular demand and peer pressure, check-in work in progress work to support | Miod Vallat | |
| the Yeelong Lemote mips-based netbook. Kernel bits only for now, needs polishing; most of this work done during h2k9 last month, although the porting effort started earlier this year. | |||
| 2009-12-11 | By popular demand and peer pressure, check-in work in progress work to support | Miod Vallat | |
| the Yeelong Lemote mips-based netbook. Kernel bits only for now, needs polishing; most of this work done during h2k9 last month, although the porting effort started earlier this year. | |||
| 2009-12-11 | More changes to avoid truncating 64 bit pointers to 32 bits and sign-extending | Miod Vallat | |
| them; allows userland with stack beyond 2GB to run. | |||
| 2009-12-10 | ansi | Theo de Raadt | |
| 2009-12-10 | spacing | Theo de Raadt | |
| 2009-12-09 | Resubmit any pending SCSI commands when we sucessfully (re)connect to | Mark Kettenis | |
| a vDisk server. This makes OpenBSD running in a guest domain survive a reboot of a control domain or service domain. | |||
| 2009-12-09 | this does not even compile | Theo de Raadt | |
| 2009-12-09 | add uhts(4). | Matthieu Herrb | |
| 2009-12-09 | By popular demand and peer pressure, check-in work in progress work to support | Miod Vallat | |
| the Yeelong Lemote mips-based netbook. Kernel bits only for now, needs polishing; most of this work done during h2k9 last month, although the porting effort started earlier this year. | |||
| 2009-12-09 | By popular demand and peer pressure, check-in work in progress work to support | Miod Vallat | |
| the Yeelong Lemote mips-based netbook. Kernel bits only for now, needs polishing; most of this work done during h2k9 last month, although the porting effort started earlier this year. | |||
| 2009-12-09 | sloppy oga, tsk tsk, you are scaring us | Theo de Raadt | |
| spotted by ckuethe and must be in immediately before my build gets there.. | |||
| 2009-12-09 | Don't try to shove I/O down the LDC if we're disconnected from the virtual | Mark Kettenis | |
| disk server. | |||
| 2009-12-09 | Remove the clean gdt bit and leave the idt part in. | Paul Irofti | |
| Fixes most laptops out there on resume. Okay deraadt@. | |||
| 2009-12-09 | add two new MD only pmap apis to amd64 and i386 (not to be used in MI | Owain Ainsworth | |
| code): pmap_flush_cache(vaddr_t, vsize_t) and pmap_flush_page(paddr_t) to flush the cache for virtual addresses and physical pages respectively using the clflush instruction. These apis will shortly be used by the agp bus_dma functions to avoid doing a wbinvd on each dmamap_sync. ok kettenis@, some comments from miod@ | |||
| 2009-12-09 | add cpufunc functions for the clflush instruction and the mfence | Owain Ainsworth | |
| instruction. ok kettenis@ as part of a larger diff. | |||
| 2009-12-09 | Detect the cache line size for the clflush instruction when we identify | Owain Ainsworth | |
| the cpu. ok kettenis@ as part of a larger diff. | |||
| 2009-12-08 | Unconditionnaly move kernel virtual memory space to XKSEG, now that previous | Miod Vallat | |
| context.S fixes allows these settings to work for kernels linked in CKSEG0. | |||
| 2009-12-08 | In cpu_switchto(), correctly ignore CKSEG0 curprocpaddr (and do not update the | Miod Vallat | |
| wired TLB entries). | |||
| 2009-12-08 | Use a whole page for msgbuf if page size is larger than 4KB, instead of | Miod Vallat | |
| the current 8KB. | |||
| 2009-12-08 | Passing kva + size as the third argument to uvm_km_free() is a bad idea. It | Mark Kettenis | |
| unmaps quite a bit more than intended, which causes weird things like watchdog resets. | |||
| 2009-12-08 | Base the revision string in the INQUIRY info on the version of the VIO protocol | Mark Kettenis | |
| used by the virtual disk server. | |||
| 2009-12-07 | Correctly skip available memory regions without (kernel) page aligned | Miod Vallat | |
| boundaries, entirely contained within a kernel page; issue raised by drahn@ | |||
| 2009-12-07 | Sync VM_MIN_ADDRESS with __LDPGSZ again. | Miod Vallat | |
| 2009-12-07 | Support for 16KB page size kernels; page size is now set in <machine/param.h> | Miod Vallat | |
| rather than <mips64/param.h>. For now, kernels are kept at 4KB to give people some time to build 16KB compatible binaries; this will change before the end of this release cycle. Use of 16KB page size kernels yields a 18% speedup (which, offset by the 1.6% slowdown caused by the pmap changes, yields a 16.6% overall speedup). | |||
| 2009-12-07 | If UPAGES would turn out to be 1, instead of using wired TLB entries to map | Miod Vallat | |
| the u area after each context switch, simply convert it to an XKPHYS address, so that access to it never faults. Idea from NetBSD. | |||
| 2009-12-07 | Use a pool to manage pmap pte pages and top level segment table, instead of | Miod Vallat | |
| directly allocating pages from uvm; this will allow us to eventually use a different kernel page size without having to alter the pmap structures layout. No functional change; measured slowdown of 1.6% for 4KB page kernels. | |||
| 2009-12-07 | When converting ARCBios memory spaces from ARCBios page size to the kernel | Miod Vallat | |
| page size, be sure to pick the strictest interval so as not to incorrectly claim ARCBios reserved data not (kernel) page size aligned as free memory. No functional change since the kernel uses the same page size as ARCBios (at the moment). | |||
| 2009-12-07 | Be sure to always check tlbp return values for strictly negative values | Miod Vallat | |
| to consider the probe as failed; tlb pair #0 does not deserve to be magic in any way. | |||
| 2009-12-07 | When netbooting the kernel image from ARCBios, our load address may not be | Miod Vallat | |
| honoured, and we may end up loaded at an uncached address (e.g. CKSEG1 on O2). This would cause the kernel to run with instruction cache disabled. Work around this by simply jumping to the address we really want to run at, at the beginning of locore. | |||
| 2009-12-07 | When trying to return to ARCBios on a 32 bit ARCBios machine (such as the O2), | Miod Vallat | |
| disable interrupts and unconditionnaly use proc0 u area as the stack, so that once ARCBios loses the upper 32 bits of the stack pointer, accessing the stack does not fault (proc0 u area is allocated in CKSEG0 for this reason). | |||
| 2009-12-04 | Send `* pwr d' instead of `pwr d' on powerdown, to have L1 broadcast the | Miod Vallat | |
| request to other bricks connected to our node. | |||
| 2009-12-04 | ahc fits again | Theo de Raadt | |
| 2009-12-03 | Stop using the linker script. Something is arranging memory just so | Theo de Raadt | |
| resulting in kernels starting up the other cpu's incorrectly on processors with very long pipelines, perhaps some bug instruction prefetch dependency in the mptramp. | |||
| 2009-12-03 | RM7000 processors with 64 TLB pairs instead of the usual 48 use an external | Miod Vallat | |
| control bit to enabled use of the extra 16, in order to be able to be used as drop-in R5000 replacement without modifying RM7000-unaware software. Because of this, when a 64 TLB RM7000 processor is detected, check that the upper 16 really will be used before deciding how many TLB the kernel needs to invalidate whe ASID wraps. | |||
| 2009-12-02 | no firmware on ramdisks; ok kettenis | Theo de Raadt | |
| 2009-12-02 | IPL_IPI priority raised to IPL_IPI > IPL_HIGH for handling TLB shootdown. | Takuya ASADA | |
| ok miod@ | |||
| 2009-12-02 | use kstack for curprocpaddr instead of proc0paddr. | Takuya ASADA | |
| ok miod@ | |||
| 2009-12-01 | Fix the mixer interface for the CE4231 sound card found in Sun machines. Still a | Edd Barrett | |
| lot of room for improvement in the driver, but atleast it works now. OK kettenis@, ratchov@ | |||
 |
index : src | |
| OpenBSD base system |
| summaryrefslogtreecommitdiff |