wireguard-openbsd/sys/arch/sgi/include/bus.h, branch jd/queueboosts WireGuard implementation for the OpenBSD kernel https://git.zx2c4.com/wireguard-openbsd/atom/sys/arch/sgi/include/bus.h?h=jd%2Fqueueboosts 2020-04-14T17:35:28Z Fix some of the more esoteric bus_space functions. The 2020-04-14T17:35:28Z kettenis kettenis@openbsd.org 2020-04-14T17:35:28Z urn:sha1:30c4e479f31e146a25569b6eda3b7bfaba926995 bus_space_read_region_n, bus_space_write_region_n and bus_space_set_region_n functions were all broken. Same fix as arm64; Thanks to patrick@ for noting that mips64 had the same code. ok visa@ The device_to_pa routine really isn't needed. We always have physical 2017-05-11T15:47:45Z visa visa@openbsd.org 2017-05-11T15:47:45Z urn:sha1:e48a128c9d41336820fcba16cc8cdc5530cdf674 addresses. While there, pave the way for BUS_DMA_64BIT (not working yet). Diff from miod@; OK dlg@ add a BUS_DMA_64BIT flag to bus_dma on all our archs. 2017-05-08T00:27:45Z dlg dlg@openbsd.org 2017-05-08T00:27:45Z urn:sha1:b0002153e68e19a4277a3f21e84cea855054f754 this is so drivers can advertise that they can handle 64 dma addresses to the platform. it may choose to handle dmamaps differently based on this flag. tweaks and ok tom@ ok kettenis@ Support BUS_DMA_NOCACHE in bus_dma(9). Memory allocations done with 2014-03-10T21:32:15Z miod miod@openbsd.org 2014-03-10T21:32:15Z urn:sha1:acc42d188fae858bccd1c6c5f0a92664952f7cc4 BUS_DMA_NOCACHE (or BUS_DMA_COHERENT if the platform does not have coherent caches) will use PMAP_NOCACHE when invoking pmap_enter(), to avoid creating cached mappings, and then evicting them from the cache. No longer allow bus_space_barrier() to be a NULL pointer on a given 2012-03-15T18:52:55Z miod miod@openbsd.org 2012-03-15T18:52:55Z urn:sha1:b15e547f0e79c7d4f23ccd40756384f1562255fc bus_space_tag on sgi, but rather always provide at least a dummy asm("sync") flavour. Saves a function pointer test at runtime. Implement bus_space_barrier() on sgi; on xbridge, this will also flush 2010-08-23T16:56:15Z miod miod@openbsd.org 2010-08-23T16:56:15Z urn:sha1:b4283f69e23c640f913cfbe51fbdd8c0d38197cf the pci write buffers. Fix bus_space_{read,write}_raw_region_# operation. 2010-04-04T11:24:27Z miod miod@openbsd.org 2010-04-04T11:24:27Z urn:sha1:ef063de97257daf1721669f8944498316891f78b Pass both the virtual address and the physical address of the memory range 2009-12-25T21:02:13Z miod miod@openbsd.org 2009-12-25T21:02:13Z urn:sha1:165cb00d5aea583fdd80457f57039e7aa9b7cc4c when invoking the cache functions. The physical address is needed when operating on physically-indexed caches, such as the L2 cache on Loongson processors. Preprocessor abuse makes sure that the physical address computation gets compiled out when running on a kernel compiled for virtually-indexed caches only, such as the sgi kernel. Get rid of the obsolet BUS_BARRIER_xxx constants for bus_space_barrier(), only 2009-07-30T21:39:13Z miod miod@openbsd.org 2009-07-30T21:39:13Z urn:sha1:7107682f8cd66f3de37a5950811a419ae4fc4202 provide and use BUS_SPACE_BARRIER_xxx. Make sure all platforms understand the flags argument of bus_space_map() and 2009-07-26T18:48:53Z miod miod@openbsd.org 2009-07-26T18:48:53Z urn:sha1:2ab138fd07f9495743be1d801fa5f60c606ab698 bus_space_alloc() as a bitmask of flags, and not a boolean controlling cacheability; and make sure the three MI BUS_SPACE_MAP_xxx values documented in the manual page are defined on all platforms as well.