Home | History | Annotate | Line # | Download | only in booke
booke_machdep.c revision 1.32.2.1
      1 /*	$NetBSD: booke_machdep.c,v 1.32.2.1 2021/04/03 22:28:34 thorpej Exp $	*/
      2 /*-
      3  * Copyright (c) 2010, 2011 The NetBSD Foundation, Inc.
      4  * All rights reserved.
      5  *
      6  * This code is derived from software contributed to The NetBSD Foundation
      7  * by Raytheon BBN Technologies Corp and Defense Advanced Research Projects
      8  * Agency and which was developed by Matt Thomas of 3am Software Foundry.
      9  *
     10  * This material is based upon work supported by the Defense Advanced Research
     11  * Projects Agency and Space and Naval Warfare Systems Center, Pacific, under
     12  * Contract No. N66001-09-C-2073.
     13  * Approved for Public Release, Distribution Unlimited
     14  *
     15  * Redistribution and use in source and binary forms, with or without
     16  * modification, are permitted provided that the following conditions
     17  * are met:
     18  * 1. Redistributions of source code must retain the above copyright
     19  *    notice, this list of conditions and the following disclaimer.
     20  * 2. Redistributions in binary form must reproduce the above copyright
     21  *    notice, this list of conditions and the following disclaimer in the
     22  *    documentation and/or other materials provided with the distribution.
     23  *
     24  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
     25  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
     26  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
     27  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
     28  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
     29  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
     30  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
     31  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
     32  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
     33  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
     34  * POSSIBILITY OF SUCH DAMAGE.
     35  */
     36 
     37 #define	__INTR_PRIVATE
     38 #define	_POWERPC_BUS_DMA_PRIVATE
     39 
     40 #include <sys/cdefs.h>
     41 __KERNEL_RCSID(0, "$NetBSD: booke_machdep.c,v 1.32.2.1 2021/04/03 22:28:34 thorpej Exp $");
     42 
     43 #include "ksyms.h"
     44 
     45 #ifdef _KERNEL_OPT
     46 #include "opt_ddb.h"
     47 #include "opt_modular.h"
     48 #include "opt_multiprocessor.h"
     49 #endif
     50 
     51 #include <sys/param.h>
     52 #include <sys/cpu.h>
     53 #include <sys/device.h>
     54 #include <sys/intr.h>
     55 #include <sys/mount.h>
     56 #include <sys/msgbuf.h>
     57 #include <sys/kernel.h>
     58 #include <sys/reboot.h>
     59 #include <sys/bus.h>
     60 #include <sys/cpu.h>
     61 
     62 #include <uvm/uvm_extern.h>
     63 
     64 #include <dev/cons.h>
     65 
     66 #include <powerpc/pcb.h>
     67 #include <powerpc/spr.h>
     68 #include <powerpc/booke/spr.h>
     69 #include <powerpc/booke/cpuvar.h>
     70 
     71 /*
     72  * Global variables used here and there
     73  */
     74 paddr_t msgbuf_paddr;
     75 psize_t pmemsize;
     76 struct vm_map *phys_map;
     77 
     78 #ifdef MODULAR
     79 register_t cpu_psluserset = PSL_USERSET;
     80 register_t cpu_pslusermod = PSL_USERMOD;
     81 register_t cpu_pslusermask = PSL_USERMASK;
     82 #endif
     83 
     84 static bus_addr_t booke_dma_phys_to_bus_mem(bus_dma_tag_t, bus_addr_t);
     85 static bus_addr_t booke_dma_bus_mem_to_phys(bus_dma_tag_t, bus_addr_t);
     86 
     87 
     88 struct powerpc_bus_dma_tag booke_bus_dma_tag = {
     89 	._dmamap_create = _bus_dmamap_create,
     90 	._dmamap_destroy = _bus_dmamap_destroy,
     91 	._dmamap_load = _bus_dmamap_load,
     92 	._dmamap_load_mbuf = _bus_dmamap_load_mbuf,
     93 	._dmamap_load_uio = _bus_dmamap_load_uio,
     94 	._dmamap_load_raw = _bus_dmamap_load_raw,
     95 	._dmamap_unload = _bus_dmamap_unload,
     96 	/*
     97 	 * The caches on BookE are coherent so we don't need to do any special
     98 	 * cache synchronization.
     99 	 */
    100 	//._dmamap_sync = _bus_dmamap_sync,
    101 	._dmamem_alloc = _bus_dmamem_alloc,
    102 	._dmamem_free = _bus_dmamem_free,
    103 	._dmamem_map = _bus_dmamem_map,
    104 	._dmamem_unmap = _bus_dmamem_unmap,
    105 	._dmamem_mmap = _bus_dmamem_mmap,
    106 	._dma_phys_to_bus_mem = booke_dma_phys_to_bus_mem,
    107 	._dma_bus_mem_to_phys = booke_dma_bus_mem_to_phys,
    108 };
    109 
    110 static bus_addr_t
    111 booke_dma_phys_to_bus_mem(bus_dma_tag_t t, bus_addr_t a)
    112 {
    113 	return a;
    114 }
    115 
    116 static bus_addr_t
    117 booke_dma_bus_mem_to_phys(bus_dma_tag_t t, bus_addr_t a)
    118 {
    119 	return a;
    120 }
    121 
    122 struct cpu_md_ops cpu_md_ops;
    123 
    124 struct cpu_softc cpu_softc[] = {
    125 	[0] = {
    126 		.cpu_ci = &cpu_info[0],
    127 	},
    128 #ifdef MULTIPROCESSOR
    129 	[CPU_MAXNUM-1] = {
    130 		.cpu_ci = &cpu_info[CPU_MAXNUM-1],
    131 	},
    132 #endif
    133 };
    134 struct cpu_info cpu_info[] = {
    135 	[0] = {
    136 		.ci_curlwp = &lwp0,
    137 		.ci_tlb_info = &pmap_tlb0_info,
    138 		.ci_softc = &cpu_softc[0],
    139 		.ci_cpl = IPL_HIGH,
    140 		.ci_idepth = -1,
    141 		.ci_pmap_kern_segtab = &pmap_kern_segtab,
    142 	},
    143 #ifdef MULTIPROCESSOR
    144 	[CPU_MAXNUM-1] = {
    145 		.ci_curlwp = NULL,
    146 		.ci_tlb_info = &pmap_tlb0_info,
    147 		.ci_softc = &cpu_softc[CPU_MAXNUM-1],
    148 		.ci_cpl = IPL_HIGH,
    149 		.ci_idepth = -1,
    150 		.ci_pmap_kern_segtab = &pmap_kern_segtab,
    151 	},
    152 #endif
    153 };
    154 __CTASSERT(__arraycount(cpu_info) == __arraycount(cpu_softc));
    155 
    156 /*
    157  * This should probably be in autoconf!				XXX
    158  */
    159 char machine[] = MACHINE;		/* from <machine/param.h> */
    160 char machine_arch[] = MACHINE_ARCH;	/* from <machine/param.h> */
    161 
    162 char bootpath[256];
    163 
    164 #if NKSYMS || defined(DDB) || defined(MODULAR)
    165 void *startsym, *endsym;
    166 #endif
    167 
    168 #if defined(MULTIPROCESSOR)
    169 volatile struct cpu_hatch_data cpu_hatch_data __cacheline_aligned;
    170 #endif
    171 
    172 int fake_mapiodev = 1;
    173 
    174 void
    175 booke_cpu_startup(const char *model)
    176 {
    177 	vaddr_t 	minaddr, maxaddr;
    178 	char 		pbuf[9];
    179 
    180 	cpu_setmodel("%s", model);
    181 
    182 	printf("%s%s", copyright, version);
    183 
    184 	format_bytes(pbuf, sizeof(pbuf), ctob((uint64_t)physmem));
    185 	printf("total memory = %s\n", pbuf);
    186 
    187 	minaddr = 0;
    188 	/*
    189 	 * Allocate a submap for physio
    190 	 */
    191 	phys_map = uvm_km_suballoc(kernel_map, &minaddr, &maxaddr,
    192 				 VM_PHYS_SIZE, 0, false, NULL);
    193 
    194 	/*
    195 	 * No need to allocate an mbuf cluster submap.  Mbuf clusters
    196 	 * are allocated via the pool allocator, and we use direct-mapped
    197 	 * pool pages.
    198 	 */
    199 
    200 	format_bytes(pbuf, sizeof(pbuf), ptoa(uvm_availmem(false)));
    201 	printf("avail memory = %s\n", pbuf);
    202 
    203 	/*
    204 	 * Register the tlb's evcnts
    205 	 */
    206 	pmap_tlb_info_evcnt_attach(curcpu()->ci_tlb_info);
    207 
    208 	/*
    209 	 * Set up the board properties database.
    210 	 */
    211 	board_info_init();
    212 
    213 	/*
    214 	 * Now that we have VM, malloc()s are OK in bus_space.
    215 	 */
    216 	bus_space_mallocok();
    217 	fake_mapiodev = 0;
    218 
    219 #ifdef MULTIPROCESSOR
    220 	pmap_kernel()->pm_active = kcpuset_running;
    221 	pmap_kernel()->pm_onproc = kcpuset_running;
    222 
    223 	for (size_t i = 1; i < __arraycount(cpu_info); i++) {
    224 		struct cpu_info * const ci = &cpu_info[i];
    225 		struct cpu_softc * const cpu = &cpu_softc[i];
    226 		cpu->cpu_ci = ci;
    227 		cpu->cpu_bst = cpu_softc[0].cpu_bst;
    228 		cpu->cpu_le_bst = cpu_softc[0].cpu_le_bst;
    229 		cpu->cpu_bsh = cpu_softc[0].cpu_bsh;
    230 		cpu->cpu_highmem = cpu_softc[0].cpu_highmem;
    231 		ci->ci_softc = cpu;
    232 		ci->ci_tlb_info = &pmap_tlb0_info;
    233 		ci->ci_cpl = IPL_HIGH;
    234 		ci->ci_idepth = -1;
    235 		ci->ci_pmap_kern_segtab = curcpu()->ci_pmap_kern_segtab;
    236 	}
    237 
    238 	kcpuset_create(&cpuset_info.cpus_running, true);
    239 	kcpuset_create(&cpuset_info.cpus_hatched, true);
    240 	kcpuset_create(&cpuset_info.cpus_paused, true);
    241 	kcpuset_create(&cpuset_info.cpus_resumed, true);
    242 	kcpuset_create(&cpuset_info.cpus_halted, true);
    243 
    244 	kcpuset_set(cpuset_info.cpus_running, cpu_number());
    245 #endif /* MULTIPROCESSOR */
    246 }
    247 
    248 static void
    249 dumpsys(void)
    250 {
    251 
    252 	printf("dumpsys: TBD\n");
    253 }
    254 
    255 /*
    256  * Halt or reboot the machine after syncing/dumping according to howto.
    257  */
    258 void
    259 cpu_reboot(int howto, char *what)
    260 {
    261 	static int syncing;
    262 	static char str[256];
    263 	char *ap = str, *ap1 = ap;
    264 
    265 	boothowto = howto;
    266 	if (!cold && !(howto & RB_NOSYNC) && !syncing) {
    267 		syncing = 1;
    268 		vfs_shutdown();		/* sync */
    269 		resettodr();		/* set wall clock */
    270 	}
    271 
    272 	splhigh();
    273 
    274 	if (!cold && (howto & RB_DUMP))
    275 		dumpsys();
    276 
    277 	doshutdownhooks();
    278 
    279 	pmf_system_shutdown(boothowto);
    280 
    281 	if ((howto & RB_POWERDOWN) == RB_POWERDOWN) {
    282 	  /* Power off here if we know how...*/
    283 	}
    284 
    285 	if (howto & RB_HALT) {
    286 		printf("The operating system has halted.\n"
    287 		    "Press any key to reboot.\n\n");
    288 		cnpollc(1);	/* For proper keyboard command handling */
    289 		cngetc();
    290 		cnpollc(0);
    291 	}
    292 
    293 	printf("rebooting\n\n");
    294 	if (what && *what) {
    295 		if (strlen(what) > sizeof str - 5)
    296 			printf("boot string too large, ignored\n");
    297 		else {
    298 			strcpy(str, what);
    299 			ap1 = ap = str + strlen(str);
    300 			*ap++ = ' ';
    301 		}
    302 	}
    303 	*ap++ = '-';
    304 	if (howto & RB_SINGLE)
    305 		*ap++ = 's';
    306 	if (howto & RB_KDB)
    307 		*ap++ = 'd';
    308 	*ap++ = 0;
    309 	if (ap[-2] == '-')
    310 		*ap1 = 0;
    311 
    312 	/* flush cache for msgbuf */
    313 	dcache_wb(msgbuf_paddr, round_page(MSGBUFSIZE));
    314 
    315 	__asm volatile("msync; isync");
    316 	(*cpu_md_ops.md_cpu_reset)();
    317 
    318 	printf("%s: md_cpu_reset() failed!\n", __func__);
    319 #ifdef DDB
    320 	for (;;)
    321 		Debugger();
    322 #else
    323 	for (;;)
    324 		/* nothing */;
    325 #endif
    326 }
    327 
    328 /*
    329  * mapiodev:
    330  *
    331  * 	Allocate vm space and mapin the I/O address. Use reserved TLB
    332  * 	mapping if one is found.
    333  */
    334 void *
    335 mapiodev(paddr_t pa, psize_t len, bool prefetchable)
    336 {
    337 	const vsize_t off = pa & PAGE_MASK;
    338 
    339 	/*
    340 	 * See if we have reserved TLB entry for the pa. This needs to be
    341 	 * true for console as we can't use uvm during early bootstrap.
    342 	 */
    343 	void * const p = tlb_mapiodev(pa, len, prefetchable);
    344 	if (p != NULL)
    345 		return p;
    346 
    347 	if (fake_mapiodev)
    348 		panic("mapiodev: no TLB entry reserved for %llx+%llx",
    349 		    (long long)pa, (long long)len);
    350 
    351 	const paddr_t orig_pa = pa;
    352 	const psize_t orig_len = len;
    353 	vsize_t align = 0;
    354 	pa = trunc_page(pa);
    355 	len = round_page(off + len);
    356 	/*
    357 	 * If we are allocating a large amount (>= 1MB) try to get an
    358 	 * aligned VA region for it so try to do a large mapping for it.
    359 	 */
    360 	if ((len & (len - 1)) == 0 && len >= 0x100000)
    361 		align = len;
    362 
    363 	vaddr_t va = uvm_km_alloc(kernel_map, len, align, UVM_KMF_VAONLY);
    364 
    365 	if (va == 0 && align > 0) {
    366 		/*
    367 		 * Large aligned request failed.  Let's just get anything.
    368 		 */
    369 		align = 0;
    370 		va = uvm_km_alloc(kernel_map, len, align, UVM_KMF_VAONLY);
    371 	}
    372 	if (va == 0)
    373 		return NULL;
    374 
    375 	if (align) {
    376 		/*
    377 		 * Now try to map that via one big TLB entry.
    378 		 */
    379 		pt_entry_t pte = pte_make_kenter_pa(pa, NULL,
    380 		    VM_PROT_READ|VM_PROT_WRITE,
    381 		    prefetchable ? 0 : PMAP_NOCACHE);
    382 		if (!tlb_ioreserve(va, len, pte)) {
    383 			void * const p0 = tlb_mapiodev(orig_pa, orig_len,
    384 			    prefetchable);
    385 			KASSERT(p0 != NULL);
    386 			return p0;
    387 		}
    388 	}
    389 
    390 	for (va += len, pa += len; len > 0; len -= PAGE_SIZE) {
    391 		va -= PAGE_SIZE;
    392 		pa -= PAGE_SIZE;
    393 		pmap_kenter_pa(va, pa, VM_PROT_READ|VM_PROT_WRITE,
    394 		    prefetchable ? 0 : PMAP_NOCACHE);
    395 	}
    396 	pmap_update(pmap_kernel());
    397 	return (void *)(va + off);
    398 }
    399 
    400 void
    401 unmapiodev(vaddr_t va, vsize_t len)
    402 {
    403 	/* Nothing to do for reserved (ie. not uvm_km_alloc'd) mappings. */
    404 	if (va < VM_MIN_KERNEL_ADDRESS || va > VM_MAX_KERNEL_ADDRESS) {
    405 		tlb_unmapiodev(va, len);
    406 		return;
    407 	}
    408 
    409 	len = round_page((va & PAGE_MASK) + len);
    410 	va = trunc_page(va);
    411 
    412 	pmap_kremove(va, len);
    413 	uvm_km_free(kernel_map, va, len, UVM_KMF_VAONLY);
    414 }
    415 
    416 void
    417 cpu_evcnt_attach(struct cpu_info *ci)
    418 {
    419 	struct cpu_softc * const cpu = ci->ci_softc;
    420 	const char * const xname = ci->ci_data.cpu_name;
    421 
    422 	evcnt_attach_dynamic_nozero(&ci->ci_ev_clock, EVCNT_TYPE_INTR,
    423 		NULL, xname, "clock");
    424 	evcnt_attach_dynamic_nozero(&cpu->cpu_ev_late_clock, EVCNT_TYPE_INTR,
    425 		NULL, xname, "late clock");
    426 	evcnt_attach_dynamic_nozero(&cpu->cpu_ev_exec_trap_sync, EVCNT_TYPE_TRAP,
    427 		NULL, xname, "exec pages synced (trap)");
    428 	evcnt_attach_dynamic_nozero(&ci->ci_ev_traps, EVCNT_TYPE_TRAP,
    429 		NULL, xname, "traps");
    430 	evcnt_attach_dynamic_nozero(&ci->ci_ev_kdsi, EVCNT_TYPE_TRAP,
    431 		&ci->ci_ev_traps, xname, "kernel DSI traps");
    432 	evcnt_attach_dynamic_nozero(&ci->ci_ev_udsi, EVCNT_TYPE_TRAP,
    433 		&ci->ci_ev_traps, xname, "user DSI traps");
    434 	evcnt_attach_dynamic_nozero(&ci->ci_ev_udsi_fatal, EVCNT_TYPE_TRAP,
    435 		&ci->ci_ev_udsi, xname, "user DSI failures");
    436 	evcnt_attach_dynamic_nozero(&ci->ci_ev_kisi, EVCNT_TYPE_TRAP,
    437 		&ci->ci_ev_traps, xname, "kernel ISI traps");
    438 	evcnt_attach_dynamic_nozero(&ci->ci_ev_isi, EVCNT_TYPE_TRAP,
    439 		&ci->ci_ev_traps, xname, "user ISI traps");
    440 	evcnt_attach_dynamic_nozero(&ci->ci_ev_isi_fatal, EVCNT_TYPE_TRAP,
    441 		&ci->ci_ev_isi, xname, "user ISI failures");
    442 	evcnt_attach_dynamic_nozero(&ci->ci_ev_scalls, EVCNT_TYPE_TRAP,
    443 		&ci->ci_ev_traps, xname, "system call traps");
    444 	evcnt_attach_dynamic_nozero(&ci->ci_ev_pgm, EVCNT_TYPE_TRAP,
    445 		&ci->ci_ev_traps, xname, "PGM traps");
    446 	evcnt_attach_dynamic_nozero(&ci->ci_ev_debug, EVCNT_TYPE_TRAP,
    447 		&ci->ci_ev_traps, xname, "debug traps");
    448 	evcnt_attach_dynamic_nozero(&ci->ci_ev_fpu, EVCNT_TYPE_TRAP,
    449 		&ci->ci_ev_traps, xname, "FPU unavailable traps");
    450 	evcnt_attach_dynamic_nozero(&ci->ci_ev_fpusw, EVCNT_TYPE_MISC,
    451 		&ci->ci_ev_fpu, xname, "FPU context switches");
    452 	evcnt_attach_dynamic_nozero(&ci->ci_ev_ali, EVCNT_TYPE_TRAP,
    453 		&ci->ci_ev_traps, xname, "user alignment traps");
    454 	evcnt_attach_dynamic_nozero(&ci->ci_ev_ali_fatal, EVCNT_TYPE_TRAP,
    455 		&ci->ci_ev_ali, xname, "user alignment traps");
    456 	evcnt_attach_dynamic_nozero(&ci->ci_ev_umchk, EVCNT_TYPE_TRAP,
    457 		&ci->ci_ev_umchk, xname, "user MCHK failures");
    458 	evcnt_attach_dynamic_nozero(&ci->ci_ev_vec, EVCNT_TYPE_TRAP,
    459 		&ci->ci_ev_traps, xname, "SPE unavailable");
    460 	evcnt_attach_dynamic_nozero(&ci->ci_ev_vecsw, EVCNT_TYPE_MISC,
    461 	    &ci->ci_ev_vec, xname, "SPE context switches");
    462 	evcnt_attach_dynamic_nozero(&ci->ci_ev_ipi, EVCNT_TYPE_INTR,
    463 		NULL, xname, "IPIs");
    464 	evcnt_attach_dynamic_nozero(&ci->ci_ev_tlbmiss_soft, EVCNT_TYPE_TRAP,
    465 		&ci->ci_ev_traps, xname, "soft tlb misses");
    466 	evcnt_attach_dynamic_nozero(&ci->ci_ev_dtlbmiss_hard, EVCNT_TYPE_TRAP,
    467 		&ci->ci_ev_traps, xname, "data tlb misses");
    468 	evcnt_attach_dynamic_nozero(&ci->ci_ev_itlbmiss_hard, EVCNT_TYPE_TRAP,
    469 		&ci->ci_ev_traps, xname, "inst tlb misses");
    470 }
    471 
    472 #ifdef MULTIPROCESSOR
    473 register_t
    474 cpu_hatch(void)
    475 {
    476 	struct cpuset_info * const csi = &cpuset_info;
    477 	const size_t id = cpu_number();
    478 
    479 	/*
    480 	 * We've hatched so tell the spinup code.
    481 	 */
    482 	kcpuset_set(csi->cpus_hatched, id);
    483 
    484 	/*
    485 	 * Loop until running bit for this cpu is set.
    486 	 */
    487 	while (!kcpuset_isset(csi->cpus_running, id)) {
    488 		continue;
    489 	}
    490 
    491 	/*
    492 	 * Now that we are active, start the clocks.
    493 	 */
    494 	cpu_initclocks();
    495 
    496 	/*
    497 	 * Return sp of the idlelwp.  Which we should be already using but ...
    498 	 */
    499 	return curcpu()->ci_curpcb->pcb_sp;
    500 }
    501 
    502 void
    503 cpu_boot_secondary_processors(void)
    504 {
    505 	volatile struct cpuset_info * const csi = &cpuset_info;
    506 	CPU_INFO_ITERATOR cii;
    507 	struct cpu_info *ci;
    508 	kcpuset_t *running;
    509 
    510 	kcpuset_create(&running, true);
    511 
    512 	for (CPU_INFO_FOREACH(cii, ci)) {
    513 		/*
    514 		 * Skip this CPU if it didn't successfully hatch.
    515 		 */
    516 		if (!kcpuset_isset(csi->cpus_hatched, cpu_index(ci)))
    517 			continue;
    518 
    519 		KASSERT(!CPU_IS_PRIMARY(ci));
    520 		KASSERT(ci->ci_data.cpu_idlelwp);
    521 
    522 		kcpuset_set(running, cpu_index(ci));
    523 	}
    524 	KASSERT(kcpuset_match(csi->cpus_hatched, running));
    525 	if (!kcpuset_iszero(running)) {
    526 		kcpuset_merge(csi->cpus_running, running);
    527 	}
    528 	kcpuset_destroy(running);
    529 }
    530 #endif
    531 
    532 uint32_t
    533 cpu_read_4(bus_addr_t a)
    534 {
    535 	struct cpu_softc * const cpu = curcpu()->ci_softc;
    536 //	printf(" %s(%p, %x, %x)", __func__, cpu->cpu_bst, cpu->cpu_bsh, a);
    537 	return bus_space_read_4(cpu->cpu_bst, cpu->cpu_bsh, a);
    538 }
    539 
    540 uint8_t
    541 cpu_read_1(bus_addr_t a)
    542 {
    543 	struct cpu_softc * const cpu = curcpu()->ci_softc;
    544 //	printf(" %s(%p, %x, %x)", __func__, cpu->cpu_bst, cpu->cpu_bsh, a);
    545 	return bus_space_read_1(cpu->cpu_bst, cpu->cpu_bsh, a);
    546 }
    547 
    548 void
    549 cpu_write_4(bus_addr_t a, uint32_t v)
    550 {
    551 	struct cpu_softc * const cpu = curcpu()->ci_softc;
    552 	bus_space_write_4(cpu->cpu_bst, cpu->cpu_bsh, a, v);
    553 }
    554 
    555 void
    556 cpu_write_1(bus_addr_t a, uint8_t v)
    557 {
    558 	struct cpu_softc * const cpu = curcpu()->ci_softc;
    559 	bus_space_write_1(cpu->cpu_bst, cpu->cpu_bsh, a, v);
    560 }
    561 
    562 void
    563 booke_sstep(struct trapframe *tf)
    564 {
    565 	uint32_t insn;
    566 
    567 	KASSERT(tf->tf_srr1 & PSL_DE);
    568 	if (ufetch_32((const void *)tf->tf_srr0, &insn) != 0)
    569 		return;
    570 
    571 	register_t dbcr0 = DBCR0_IAC1 | DBCR0_IDM;
    572 	register_t dbcr1 = DBCR1_IAC1US_USER | DBCR1_IAC1ER_DS1;
    573 	if ((insn >> 28) == 4) {
    574 		uint32_t iac2 = 0;
    575 		if ((insn >> 26) == 0x12) {
    576 			const int32_t off = (((int32_t)insn << 6) >> 6) & ~3;
    577 			iac2 = ((insn & 2) ? 0 : tf->tf_srr0) + off;
    578 			dbcr0 |= DBCR0_IAC2;
    579 		} else if ((insn >> 26) == 0x10) {
    580 			const int16_t off = insn & ~3;
    581 			iac2 = ((insn & 2) ? 0 : tf->tf_srr0) + off;
    582 			dbcr0 |= DBCR0_IAC2;
    583 		} else if ((insn & 0xfc00fffe) == 0x4c000420) {
    584 			iac2 = tf->tf_ctr;
    585 			dbcr0 |= DBCR0_IAC2;
    586 		} else if ((insn & 0xfc00fffe) == 0x4c000020) {
    587 			iac2 = tf->tf_lr;
    588 			dbcr0 |= DBCR0_IAC2;
    589 		}
    590 		if (dbcr0 & DBCR0_IAC2) {
    591 			dbcr1 |= DBCR1_IAC2US_USER | DBCR1_IAC2ER_DS1;
    592 			mtspr(SPR_IAC2, iac2);
    593 		}
    594 	}
    595 	mtspr(SPR_IAC1, tf->tf_srr0 + 4);
    596 	mtspr(SPR_DBCR1, dbcr1);
    597 	mtspr(SPR_DBCR0, dbcr0);
    598 }
    599 
    600 #ifdef DIAGNOSTIC
    601 static inline void
    602 swap_data(uint64_t *data, size_t a, size_t b)
    603 {
    604 	uint64_t swap = data[a];
    605 	data[a] = data[b];
    606 	data[b] = swap;
    607 }
    608 
    609 static void
    610 sort_data(uint64_t *data, size_t count)
    611 {
    612 #if 0
    613 	/*
    614 	 * Mostly classic bubble sort
    615 	 */
    616 	do {
    617 		size_t new_count = 0;
    618 		for (size_t i = 1; i < count; i++) {
    619 			if (tbs[i - 1] > tbs[i]) {
    620 				swap_tbs(tbs, i - 1, i);
    621 				new_count = i;
    622 			}
    623 		}
    624 		count = new_count;
    625 	} while (count > 0);
    626 #else
    627 	/*
    628 	 * Comb sort
    629 	 */
    630 	size_t gap = count;
    631 	bool swapped = false;
    632 	while (gap > 1 || swapped) {
    633 		if (gap > 1) {
    634 			/*
    635 			 * phi = (1 + sqrt(5)) / 2 [golden ratio]
    636 			 * N = 1 / (1 - e^-phi)) = 1.247330950103979
    637 			 *
    638 			 * We want to but can't use floating point to calculate
    639 			 *	gap = (size_t)((double)gap / N)
    640 			 *
    641 			 * So we will use the multicative inverse of N
    642 			 * (module 65536) to achieve the division.
    643 			 *
    644 			 * iN = 2^16 / 1.24733... = 52540
    645 			 * x / N == (x * iN) / 65536
    646 			 */
    647 			gap = (gap * 52540) / 65536;
    648 		}
    649 
    650 		swapped = false;
    651 
    652 		for (size_t i = 0; gap + i < count; i++) {
    653 			if (data[i] > data[i + gap]) {
    654 				swap_data(data, i, i + gap);
    655 				swapped = true;
    656 			}
    657 		}
    658 	}
    659 #endif
    660 }
    661 #endif
    662 
    663 void
    664 dump_splhist(struct cpu_info *ci, void (*pr)(const char *, ...))
    665 {
    666 #ifdef DIAGNOSTIC
    667 	struct cpu_softc * const cpu = ci->ci_softc;
    668 	uint64_t tbs[NIPL*NIPL];
    669 	size_t ntbs = 0;
    670 	for (size_t to = 0; to < NIPL; to++) {
    671 		for (size_t from = 0; from < NIPL; from++) {
    672 			uint64_t tb = cpu->cpu_spl_tb[to][from];
    673 			if (tb == 0)
    674 				continue;
    675 			tbs[ntbs++] = (tb << 8) | (to << 4) | from;
    676 		}
    677 	}
    678 	sort_data(tbs, ntbs);
    679 
    680 	if (pr == NULL)
    681 		pr = printf;
    682 	uint64_t last_tb = 0;
    683 	for (size_t i = 0; i < ntbs; i++) {
    684 		uint64_t tb = tbs[i];
    685 		size_t from = tb & 15;
    686 		size_t to = (tb >> 4) & 15;
    687 		tb >>= 8;
    688 		(*pr)("%s(%zu) from %zu at %"PRId64"",
    689 		     from < to ? "splraise" : "splx",
    690 		     to, from, tb);
    691 		if (last_tb && from != IPL_NONE)
    692 			(*pr)(" (+%"PRId64")", tb - last_tb);
    693 		(*pr)("\n");
    694 		last_tb = tb;
    695 	}
    696 #endif
    697 }
    698