/* $NetBSD: vm_machdep.c,v 1.102 2019/02/11 07:51:46 macallan Exp $ */ /* * Copyright (c) 1996-2002 Eduardo Horvath. All rights reserved. * Copyright (c) 1996 * The President and Fellows of Harvard College. All rights reserved. * Copyright (c) 1992, 1993 * The Regents of the University of California. All rights reserved. * * This software was developed by the Computer Systems Engineering group * at Lawrence Berkeley Laboratory under DARPA contract BG 91-66 and * contributed to Berkeley. * * All advertising materials mentioning features or use of this software * must display the following acknowledgement: * This product includes software developed by the University of * California, Lawrence Berkeley Laboratory. * This product includes software developed by Harvard University. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. All advertising materials mentioning features or use of this software * must display the following acknowledgement: * This product includes software developed by Harvard University. * This product includes software developed by the University of * California, Berkeley and its contributors. * 4. Neither the name of the University nor the names of its contributors * may be used to endorse or promote products derived from this software * without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * @(#)vm_machdep.c 8.2 (Berkeley) 9/23/93 */ #include <sys/cdefs.h> __KERNEL_RCSID(0, "$NetBSD: vm_machdep.c,v 1.102 2019/02/11 07:51:46 macallan Exp $"); #include "opt_multiprocessor.h" #include <sys/param.h> #include <sys/systm.h> #include <sys/proc.h> #include <sys/core.h> #include <sys/buf.h> #include <sys/exec.h> #include <sys/vnode.h> #include <sys/cpu.h> #include <uvm/uvm_extern.h> #include <machine/cpu.h> #include <machine/frame.h> #include <machine/pcb.h> #include <machine/trap.h> #include <sys/bus.h> /* * Map a user I/O request into kernel virtual address space. * Note: the pages are already locked by uvm_vslock(), so we * do not need to pass an access_type to pmap_enter(). */ int vmapbuf(struct buf *bp, vsize_t len) { struct pmap *upmap; vaddr_t uva; /* User VA (map from) */ vaddr_t kva; /* Kernel VA (new to) */ paddr_t pa; /* physical address */ vsize_t off; if ((bp->b_flags & B_PHYS) == 0) panic("vmapbuf"); bp->b_saveaddr = bp->b_data; uva = trunc_page((vaddr_t)bp->b_data); off = (vaddr_t)bp->b_data - uva; len = round_page(off + len); kva = uvm_km_alloc(kernel_map, len, 0, UVM_KMF_VAONLY | UVM_KMF_WAITVA); bp->b_data = (void *)(kva + off); upmap = vm_map_pmap(&bp->b_proc->p_vmspace->vm_map); do { if (pmap_extract(upmap, uva, &pa) == FALSE) panic("vmapbuf: null page frame"); /* Now map the page into kernel space. */ pmap_kenter_pa(kva, pa, VM_PROT_READ | VM_PROT_WRITE, 0); uva += PAGE_SIZE; kva += PAGE_SIZE; len -= PAGE_SIZE; } while (len); pmap_update(pmap_kernel()); return 0; } /* * Unmap a previously-mapped user I/O request. */ void vunmapbuf(struct buf *bp, vsize_t len) { vaddr_t kva; vsize_t off; if ((bp->b_flags & B_PHYS) == 0) panic("vunmapbuf"); kva = trunc_page((vaddr_t)bp->b_data); off = (vaddr_t)bp->b_data - kva; len = round_page(off + len); pmap_kremove(kva, len); uvm_km_free(kernel_map, kva, len, UVM_KMF_VAONLY); bp->b_data = bp->b_saveaddr; bp->b_saveaddr = NULL; } void cpu_proc_fork(struct proc *p1, struct proc *p2) { p2->p_md.md_flags = p1->p_md.md_flags; } /* * The offset of the topmost frame in the kernel stack. */ #ifdef __arch64__ #define TOPFRAMEOFF (USPACE-sizeof(struct trapframe)-CC64FSZ) #define STACK_OFFSET BIAS #else #undef trapframe #define trapframe trapframe64 #undef rwindow #define rwindow rwindow32 #define TOPFRAMEOFF (USPACE-sizeof(struct trapframe)-CC64FSZ) #define STACK_OFFSET 0 #endif #ifdef DEBUG char cpu_forkname[] = "cpu_lwp_fork()"; #endif /* * Finish a fork operation, with lwp l2 nearly set up. * Copy and update the pcb and trap frame, making the child ready to run. * * Rig the child's kernel stack so that it will start out in * lwp_trampoline() and call child_return() with l2 as an * argument. This causes the newly-created child process to go * directly to user level with an apparent return value of 0 from * fork(), while the parent process returns normally. * * l1 is the process being forked; if l1 == &lwp0, we are creating * a kernel thread, and the return path and argument are specified with * `func' and `arg'. * * If an alternate user-level stack is requested (with non-zero values * in both the stack and stacksize args), set up the user stack pointer * accordingly. */ void lwp_trampoline(void); void cpu_lwp_fork(register struct lwp *l1, register struct lwp *l2, void *stack, size_t stacksize, void (*func)(void *), void *arg) { struct pcb *opcb = lwp_getpcb(l1); struct pcb *npcb = lwp_getpcb(l2); struct trapframe *tf2; struct rwindow *rp; /* * Save all user registers to l1's stack or, in the case of * user registers and invalid stack pointers, to opcb. * We then copy the whole pcb to l2; when switch() selects l2 * to run, it will run at the `lwp_trampoline' stub, rather * than returning at the copying code below. * * If process l1 has an FPU state, we must copy it. If it is * the FPU user, we must save the FPU state first. */ #ifdef NOTDEF_DEBUG printf("cpu_lwp_fork()\n"); #endif if (l1 == curlwp) { write_user_windows(); /* * We're in the kernel, so we don't really care about * %ccr or %asi. We do want to duplicate %pstate and %cwp. */ opcb->pcb_pstate = getpstate(); opcb->pcb_cwp = getcwp(); } #ifdef DIAGNOSTIC else if (l1 != &lwp0) panic("cpu_lwp_fork: curlwp"); #endif #ifdef DEBUG /* prevent us from having NULL lastcall */ opcb->lastcall = cpu_forkname; #else opcb->lastcall = NULL; #endif memcpy(npcb, opcb, sizeof(struct pcb)); if (l1->l_md.md_fpstate) { fpusave_lwp(l1, true); l2->l_md.md_fpstate = pool_cache_get(fpstate_cache, PR_WAITOK); memcpy(l2->l_md.md_fpstate, l1->l_md.md_fpstate, sizeof(struct fpstate64)); } else l2->l_md.md_fpstate = NULL; /* * Setup (kernel) stack frame that will by-pass the child * out of the kernel. (The trap frame invariably resides at * the tippity-top of the u. area.) */ tf2 = l2->l_md.md_tf = (struct trapframe *) ((long)npcb + USPACE - sizeof(*tf2)); /* Copy parent's trapframe */ *tf2 = *(struct trapframe *)((long)opcb + USPACE - sizeof(*tf2)); /* * If specified, give the child a different stack. */ if (stack != NULL) tf2->tf_out[6] = (uint64_t)(u_long)stack + stacksize; /* * Need to create a %tstate if we are forking our first userland * process - in all other cases we inherit from the parent. */ if (l2->l_proc->p_pid == 1) tf2->tf_tstate = (ASI_PRIMARY_NO_FAULT<<TSTATE_ASI_SHIFT) | ((PSTATE_USER)<<TSTATE_PSTATE_SHIFT); /* * Set return values in child mode and clear condition code, * in case we end up running a signal handler before returning * to userland. */ tf2->tf_out[0] = 0; tf2->tf_out[1] = 1; tf2->tf_tstate &= ~TSTATE_CCR; /* Construct kernel frame to return to in cpu_switch() */ rp = (struct rwindow *)((u_long)npcb + TOPFRAMEOFF); *rp = *(struct rwindow *)((u_long)opcb + TOPFRAMEOFF); rp->rw_local[0] = (long)func; /* Function to call */ rp->rw_local[1] = (long)arg; /* and its argument */ rp->rw_local[2] = (long)l2; /* new lwp */ npcb->pcb_pc = (long)lwp_trampoline - 8; npcb->pcb_sp = (long)rp - STACK_OFFSET; } static inline void fpusave_cpu(bool save) { struct lwp *l = fplwp; if (l == NULL) return; if (save) savefpstate(l->l_md.md_fpstate); else clearfpstate(); fplwp = NULL; } void fpusave_lwp(struct lwp *l, bool save) { #ifdef MULTIPROCESSOR volatile struct cpu_info *ci; if (l == fplwp) { int s = intr_disable(); fpusave_cpu(save); intr_restore(s); return; } for (ci = cpus; ci != NULL; ci = ci->ci_next) { int spincount, retrycount=0; if (ci == curcpu() || !CPUSET_HAS(cpus_active, ci->ci_index)) continue; if (ci->ci_fplwp != l) continue; again: sparc64_send_ipi(ci->ci_cpuid, save ? sparc64_ipi_save_fpstate : sparc64_ipi_drop_fpstate, (uintptr_t)l, 0); spincount = 0; while (ci->ci_fplwp == l) { membar_Sync(); spincount++; if (spincount > 10000000) { printf("fpusave_lwp ipi didn't (%d)\n", retrycount); retrycount++; goto again; } } if (retrycount > 0) printf("spincount %d\n", spincount); break; } #else if (l == fplwp) fpusave_cpu(save); #endif } void cpu_lwp_free(struct lwp *l, int proc) { if (l->l_md.md_fpstate != NULL) fpusave_lwp(l, false); } void cpu_lwp_free2(struct lwp *l) { struct fpstate64 *fs; if ((fs = l->l_md.md_fpstate) != NULL) pool_cache_put(fpstate_cache, fs); } int cpu_lwp_setprivate(lwp_t *l, void *addr) { struct trapframe *tf = l->l_md.md_tf; tf->tf_global[7] = (uintptr_t)addr; return 0; }
