diff options
author | Per Fogelstrom <pefo@cvs.openbsd.org> | 2004-09-23 08:42:39 +0000 |
---|---|---|
committer | Per Fogelstrom <pefo@cvs.openbsd.org> | 2004-09-23 08:42:39 +0000 |
commit | e083531d6f2e128a1b8903afe2436350fa7818b5 (patch) | |
tree | d867c1d8de66cde58b1bc5ce6b63e96a182cb8df /sys/arch/mips64 | |
parent | c7855bd25894126d167a659dfd49d90d6725fd8d (diff) |
More pmap bugs cleaned out. Some old, some new.
Better structured pmap.c.
Evil still in there, more work to do.
Diffstat (limited to 'sys/arch/mips64')
-rw-r--r-- | sys/arch/mips64/include/pmap.h | 7 | ||||
-rw-r--r-- | sys/arch/mips64/mips64/pmap.c | 823 | ||||
-rw-r--r-- | sys/arch/mips64/mips64/trap.c | 96 |
3 files changed, 450 insertions, 476 deletions
diff --git a/sys/arch/mips64/include/pmap.h b/sys/arch/mips64/include/pmap.h index a67011ac8db..6514b8831dd 100644 --- a/sys/arch/mips64/include/pmap.h +++ b/sys/arch/mips64/include/pmap.h @@ -1,4 +1,4 @@ -/* $OpenBSD: pmap.h,v 1.4 2004/09/17 19:19:05 miod Exp $ */ +/* $OpenBSD: pmap.h,v 1.5 2004/09/23 08:42:38 pefo Exp $ */ /* * Copyright (c) 1987 Carnegie-Mellon University @@ -93,11 +93,12 @@ extern struct pmap kernel_pmap_store; #define pmap_resident_count(pmap) ((pmap)->pm_stats.resident_count) #define pmap_wired_count(pmap) ((pmap)->pm_stats.wired_count) -#define pmap_kernel() (&kernel_pmap_store) +#define pmap_kernel() (&kernel_pmap_store) +#define pmap_phys_address(ppn) ptoa(ppn) #define PMAP_STEAL_MEMORY /* Enable 'stealing' during boot */ -#define PMAP_PREFER(pa, va) pmap_prefer(pa, va) +#define PMAP_PREFER(pa, va) pmap_prefer(pa, va) #define pmap_update(x) /* nothing */ diff --git a/sys/arch/mips64/mips64/pmap.c b/sys/arch/mips64/mips64/pmap.c index f1bc75b0650..6847a45e387 100644 --- a/sys/arch/mips64/mips64/pmap.c +++ b/sys/arch/mips64/mips64/pmap.c @@ -1,4 +1,4 @@ -/* $OpenBSD: pmap.c,v 1.11 2004/09/22 07:27:22 miod Exp $ */ +/* $OpenBSD: pmap.c,v 1.12 2004/09/23 08:42:38 pefo Exp $ */ /* * Copyright (c) 2001-2004 Opsycon AB (www.opsycon.se / www.opsycon.com) @@ -38,20 +38,23 @@ #include <sys/shm.h> #endif -#include <uvm/uvm.h> - #include <machine/pte.h> #include <machine/cpu.h> #include <machine/autoconf.h> #include <machine/memconf.h> +#include <machine/vmparam.h> #include <mips64/archtype.h> +#include <uvm/uvm.h> + extern void mem_zero_page(vaddr_t); +/* flags for pv_entry */ #define PV_UNCACHED 0x0001 /* Page is mapped unchached */ #define PV_CACHED 0x0002 /* Page has been cached */ #define PV_ATTR_MOD 0x0004 #define PV_ATTR_REF 0x0008 +#define PV_PRESERVE (PV_UNCACHED|PV_CACHED|PV_ATTR_MOD|PV_ATTR_REF) struct pool pmap_pmap_pool; struct pool pmap_pv_pool; @@ -65,6 +68,7 @@ struct pool pmap_pv_pool; int pmap_pv_lowat = PMAP_PV_LOWAT; int pmap_alloc_tlbpid(struct proc *); +int pmap_enter_pv(pmap_t, vaddr_t, vm_page_t, u_int *); int pmap_page_alloc(vaddr_t *); void pmap_page_cache(vm_page_t, int); void pmap_remove_pv(pmap_t, vaddr_t, paddr_t); @@ -94,8 +98,6 @@ struct { int pvsearch; } remove_stats; -#define stat_count(what) (what)++ - #define PDB_FOLLOW 0x0001 #define PDB_INIT 0x0002 #define PDB_ENTER 0x0004 @@ -111,7 +113,18 @@ struct { #define PDB_WIRING 0x4000 #define PDB_PVDUMP 0x8000 +#define DPRINTF(flag, printdata) \ + if (pmapdebug & (flag)) \ + printf printdata; + +#define stat_count(what) (what)++ int pmapdebug = 0x0; + +#else + +#define DPRINTF(flag, printdata) +#define stat_count(what) + #endif /* PMAPDEBUG */ @@ -245,11 +258,7 @@ void pmap_init() { -#ifdef PMAPDEBUG - if (pmapdebug & (PDB_FOLLOW|PDB_INIT)) { - printf("pmap_init()\n"); - } -#endif + DPRINTF(PDB_FOLLOW|PDB_INIT, ("pmap_init()\n")); #if 0 /* too early */ pool_setlowat(&pmap_pv_pool, pmap_pv_lowat); @@ -263,30 +272,6 @@ pg_to_pvh(struct vm_page *pg) return &pg->mdpage.pvent; } -int -pmap_page_alloc(vaddr_t *ret) -{ - vm_page_t pg; - pv_entry_t pv; - vaddr_t va; - - pg = uvm_pagealloc(NULL, 0, NULL, UVM_PGA_USERESERVE | UVM_PGA_ZERO); - if (pg == NULL) - return ENOMEM; - - pv = pg_to_pvh(pg); - va = PHYS_TO_KSEG0(VM_PAGE_TO_PHYS(pg)); - if ((pv->pv_flags & PV_CACHED) && - ((pv->pv_va ^ va) & CpuCacheAliasMask) != 0) { - Mips_SyncDCachePage(pv->pv_va); - } - pv->pv_va = va; - pv->pv_flags = PV_CACHED; - - *ret = va; - return 0; -} - /* * Create and return a physical map. */ @@ -298,11 +283,7 @@ pmap_create() extern struct vmspace vmspace0; extern struct user *proc0paddr; -#ifdef PMAPDEBUG - if (pmapdebug & (PDB_FOLLOW|PDB_CREATE)) { - printf("pmap_create()\n"); - } -#endif + DPRINTF(PDB_FOLLOW|PDB_CREATE, ("pmap_create()\n")); pmap = pool_get(&pmap_pmap_pool, PR_WAITOK); bzero(pmap, sizeof(*pmap)); @@ -359,11 +340,7 @@ pmap_destroy(pmap) { int count; -#ifdef PMAPDEBUG - if (pmapdebug & (PDB_FOLLOW|PDB_CREATE)) { - printf("pmap_destroy(%x)\n", pmap); - } -#endif + DPRINTF(PDB_FOLLOW|PDB_CREATE, ("pmap_destroy(%x)\n", pmap)); simple_lock(&pmap->pm_lock); count = --pmap->pm_count; @@ -386,7 +363,7 @@ pmap_destroy(pmap) #ifdef PARANOIA for (j = 0; j < NPTEPG; j++) { if ((pte+j)->pt_entry) - panic("pmap_release: segmap not empty"); + panic("pmap_destroy: segmap not empty"); } #endif Mips_HitInvalidateDCache((vaddr_t)pte, PAGE_SIZE); @@ -409,11 +386,8 @@ pmap_reference(pmap) pmap_t pmap; { -#ifdef PMAPDEBUG - if (pmapdebug & PDB_FOLLOW) { - printf("pmap_reference(%x)\n", pmap); - } -#endif + DPRINTF(PDB_FOLLOW, ("pmap_reference(%x)\n", pmap)); + if (pmap) { simple_lock(&pmap->pm_lock); pmap->pm_count++; @@ -460,12 +434,10 @@ pmap_remove(pmap, sva, eva) pt_entry_t *pte; unsigned entry; -#ifdef PMAPDEBUG - if (pmapdebug & (PDB_FOLLOW|PDB_REMOVE|PDB_PROTECT)) { - printf("pmap_remove(%x, %x, %x)\n", pmap, sva, eva); - } + DPRINTF(PDB_FOLLOW|PDB_REMOVE|PDB_PROTECT, + ("pmap_remove(%x, %x, %x)\n", pmap, sva, eva)); + stat_count(remove_stats.calls); -#endif if (pmap == NULL) { return; } @@ -488,11 +460,9 @@ pmap_remove(pmap, sva, eva) /* * Flush the TLB for the given address. */ - pte->pt_entry = PG_NV | PG_G; /* See above about G bit */ + pte->pt_entry = PG_NV | PG_G; tlb_flush_addr(sva); -#ifdef PMAPDEBUG stat_count(remove_stats.flushes); -#endif } return; } @@ -532,9 +502,7 @@ pmap_remove(pmap, sva, eva) if (pmap->pm_tlbgen == tlbpid_gen) { tlb_flush_addr(sva | (pmap->pm_tlbpid << VMTLB_PID_SHIFT)); -#ifdef PMAPDEBUG stat_count(remove_stats.flushes); -#endif } } } @@ -553,12 +521,13 @@ pmap_page_protect(pg, prot) pv_entry_t pv; vaddr_t va; -#ifdef PMAPDEBUG - if ((pmapdebug & (PDB_FOLLOW|PDB_PROTECT)) || - ((prot == VM_PROT_NONE) && (pmapdebug & PDB_REMOVE))) { - printf("pmap_page_protect(%x, %x)\n", pg, prot); + if (prot == VM_PROT_NONE) { + DPRINTF(PDB_REMOVE, ("pmap_page_protect(%p, %p)\n", pg, prot)); + } else { + DPRINTF(PDB_FOLLOW|PDB_PROTECT, + ("pmap_page_protect(%p, %p)\n", pg, prot)); } -#endif + if (pg == NULL) return; @@ -608,11 +577,9 @@ pmap_protect(pmap, sva, eva, prot) unsigned entry; u_int p; -#ifdef PMAPDEBUG - if (pmapdebug & (PDB_FOLLOW|PDB_PROTECT)) { - printf("pmap_protect(%x, %x, %x, %x)\n", pmap, sva, eva, prot); - } -#endif + DPRINTF(PDB_FOLLOW|PDB_PROTECT, + ("pmap_protect(%p, %p, %p, %p)\n", pmap, sva, eva, prot)); + if (pmap == NULL) return; @@ -688,71 +655,6 @@ pmap_protect(pmap, sva, eva, prot) } /* - * Return RO protection of page. - */ -int -pmap_is_page_ro(pmap, va, entry) - pmap_t pmap; - vaddr_t va; - int entry; -{ - return(entry & PG_RO); -} - -/* - * pmap_page_cache: - * - * Change all mappings of a page to cached/uncached. - */ -void -pmap_page_cache(pg,mode) - vm_page_t pg; - int mode; -{ - pv_entry_t pv; - pt_entry_t *pte; - unsigned entry; - unsigned newmode; - -#ifdef PMAPDEBUG - if (pmapdebug & (PDB_FOLLOW|PDB_ENTER)) { - printf("pmap_page_uncache(%x)\n", pa); - } -#endif - newmode = mode & PV_UNCACHED ? PG_UNCACHED : PG_CACHED; - pv = pg_to_pvh(pg); - while (pv) { - pv->pv_flags = (pv->pv_flags & ~PV_UNCACHED) | mode; - if (pv->pv_pmap == pmap_kernel()) { - /* - * Change entries in kernel pmap. - */ - pte = kvtopte(pv->pv_va); - entry = pte->pt_entry; - if (entry & PG_V) { - entry = (entry & ~PG_CACHEMODE) | newmode; - pte->pt_entry = entry; - tlb_update(pv->pv_va, entry); - } - } - else { - if ((pte = pmap_segmap(pv->pv_pmap, pv->pv_va))) { - pte += uvtopte(pv->pv_va); - entry = pte->pt_entry; - if (entry & PG_V) { - entry = (entry & ~PG_CACHEMODE) | newmode; - pte->pt_entry = entry; - if (pv->pv_pmap->pm_tlbgen == tlbpid_gen) - tlb_update(pv->pv_va | (pv->pv_pmap->pm_tlbpid << - VMTLB_PID_SHIFT), entry); - } - } - } - pv = pv->pv_next; - } -} - -/* * Insert the given physical page (p) at * the specified virtual address (v) in the * target physical map with the protection requested. @@ -773,23 +675,16 @@ pmap_enter(pmap, va, pa, prot, flags) u_int npte; vm_page_t pg; -#ifdef PMAPDEBUG - if (pmapdebug & (PDB_FOLLOW|PDB_ENTER)) { - printf("pmap_enter(%x, %x, %x, %x, %x)\n", - pmap, va, pa, prot, stat); - } -#endif + DPRINTF(PDB_FOLLOW|PDB_ENTER, + ("pmap_enter(%p, %p, %p, %p, %p)\n", pmap, va, pa, prot, flags)); + #ifdef DIAGNOSTIC if (pmap == pmap_kernel()) { -#ifdef PMAPDEBUG - enter_stats.kernel++; -#endif + stat_count(enter_stats.kernel); if (va < VM_MIN_KERNEL_ADDRESS || va >= virtual_end) panic("pmap_enter: kva %p", va); } else { -#ifdef PMAPDEBUG - enter_stats.user++; -#endif + stat_count(enter_stats.user); if (va >= VM_MAXUSER_ADDRESS) panic("pmap_enter: uva %p", va); } @@ -798,24 +693,21 @@ pmap_enter(pmap, va, pa, prot, flags) pg = PHYS_TO_VM_PAGE(pa); if (pg != NULL) { - pv_entry_t pv, npv; + pv_entry_t pv; pv = pg_to_pvh(pg); if (!(prot & VM_PROT_WRITE)) { npte = PG_ROPAGE; } else { - if ((int)va < 0) { + if ((int64_t)va < 0) { /* * Don't bother to trap on kernel writes, * just record page as dirty. */ npte = PG_RWPAGE; -#if 0 /*XXX*/ - pg->flags &= ~PG_CLEAN; -#endif } else { - if (!(pg->flags & PG_CLEAN)) { + if (pv->pv_flags & PV_ATTR_MOD) { npte = PG_RWPAGE; } else { npte = PG_CWPAGE; @@ -823,112 +715,19 @@ pmap_enter(pmap, va, pa, prot, flags) } } -#ifdef PMAPDEBUG - stat_count(enter_stats.managed); -#endif - /* - * Enter the pmap and virtual address into the - * physical to virtual map table. - */ - -#ifdef PMAPDEBUG - if (pmapdebug & PDB_ENTER) { - printf("pmap_enter: pv %x: was %x/%x/%x\n", - pv, pv->pv_va, pv->pv_pmap, pv->pv_next); - } -#endif - - if (pv->pv_pmap == NULL) { - /* - * No entries yet, use header as the first entry - */ - -#ifdef PMAPDEBUG - if (pmapdebug & PDB_PVENTRY) { - printf("pmap_enter: first pv: pmap %x va %x pa %p\n", - pmap, va, pa); - } - stat_count(enter_stats.firstpv); -#endif - - Mips_SyncDCachePage(pv->pv_va); - - pv->pv_va = va; - pv->pv_flags = PV_CACHED; - pv->pv_pmap = pmap; - pv->pv_next = NULL; - } else { - if (pv->pv_flags & PV_UNCACHED) { - npte = (npte & ~PG_CACHEMODE) | PG_UNCACHED; - } else if (CpuCacheAliasMask != 0) { - /* - * There is at least one other VA mapping this page. - * Check if they are cache index compatible. If not - * remove all mappings, flush the cache and set page - * to be mapped uncached. Caching will be restored - * when pages are mapped compatible again. NOT! - */ - for (npv = pv; npv; npv = npv->pv_next) { - /* - * Check cache aliasing incompatibility - */ - if (((npv->pv_va ^ va) & CpuCacheAliasMask) != 0) { - printf("pmap_enter: uncached mapping for pa %p, va %p != %p.\n", pa, npv->pv_va, va); - pmap_page_cache(pg,PV_UNCACHED); - Mips_SyncCache(); - pv->pv_flags &= ~PV_CACHED; - npte = (npte & ~PG_CACHEMODE) | PG_UNCACHED; - break; - } - } - } + /* Set page referenced/modified status based on flags */ + if (flags & VM_PROT_WRITE) + pv->pv_flags |= PV_ATTR_MOD | PV_ATTR_REF; + else if (flags & VM_PROT_ALL) + pv->pv_flags |= PV_ATTR_REF; - /* - * There is at least one other VA mapping this page. - * Place this entry after the header. - * - * Note: the entry may already be in the table if - * we are only changing the protection bits. - */ - for (npv = pv; npv; npv = npv->pv_next) { - if (pmap == npv->pv_pmap && va == npv->pv_va) { - goto fnd; - } - } - -#ifdef PMAPDEBUG - if (pmapdebug & PDB_PVENTRY) { - printf("pmap_enter: new pv: pmap %x va %x pa %p\n", - pmap, va, pa); - } -#endif - - /* can this cause us to recurse forever? */ - npv = pmap_pv_alloc(); - if (npv == NULL) { - panic("pmap_pv_alloc() failed"); - } - npv->pv_va = va; - npv->pv_pmap = pmap; - npv->pv_next = pv->pv_next; - npv->pv_flags = pv->pv_flags; - pv->pv_next = npv; - -#ifdef PMAPDEBUG - if (!npv->pv_next) - stat_count(enter_stats.secondpv); -#endif - fnd: - ; - } + stat_count(enter_stats.managed); } else { /* * Assumption: if it is not part of our managed memory - * then it must be device memory which may be volitile. + * then it must be device memory which may be volatile. */ -#ifdef PMAPDEBUG stat_count(enter_stats.unmanaged); -#endif if (prot & VM_PROT_WRITE) { npte = PG_IOPAGE & ~PG_G; } else { @@ -937,16 +736,22 @@ pmap_enter(pmap, va, pa, prot, flags) } if (pmap == pmap_kernel()) { + if (pg != NULL) { + if (pmap_enter_pv(pmap, va, pg, &npte) != 0) { + if (flags & PMAP_CANFAIL) + return ENOMEM; + panic("pmap_enter: pmap_enter_pv() failed"); + } + } + pte = kvtopte(va); npte |= vad_to_pfn(pa) | PG_ROPAGE | PG_G; if (!(pte->pt_entry & PG_V)) { pmap->pm_stats.resident_count++; } - if (pa != pfn_to_pad(pte->pt_entry)) { + if ((pte->pt_entry & PG_V) && pa != pfn_to_pad(pte->pt_entry)) { pmap_remove(pmap, va, va + NBPG); -#ifdef PMAPDEBUG stat_count(enter_stats.mchange); -#endif } /* @@ -954,7 +759,7 @@ pmap_enter(pmap, va, pa, prot, flags) */ pte->pt_entry = npte; tlb_update(va, npte); - return (KERN_SUCCESS); + return 0; } /* @@ -972,6 +777,14 @@ pmap_enter(pmap, va, pa, prot, flags) pmap_segmap(pmap, va) = pte = (pt_entry_t *)nva; } + if (pg != NULL) { + if (pmap_enter_pv(pmap, va, pg, &npte) != 0) { + if (flags & PMAP_CANFAIL) + return ENOMEM; + panic("pmap_enter: pmap_enter_pv() failed"); + } + } + pte += uvtopte(va); /* @@ -980,21 +793,16 @@ pmap_enter(pmap, va, pa, prot, flags) * MIPS pages in a OpenBSD page. */ npte |= vad_to_pfn(pa); - -#ifdef PMAPDEBUG - if (pmapdebug & PDB_ENTER) { - printf("pmap_enter: new pte %x", npte); - if (pmap->pm_tlbgen == tlbpid_gen) - printf(" tlbpid %d", pmap->pm_tlbpid); - printf("\n"); + if (pmap->pm_tlbgen == tlbpid_gen) { + DPRINTF(PDB_ENTER, ("pmap_enter: new pte %x tlbpid %d\n", + npte, pmap->pm_tlbpid)); + } else { + DPRINTF(PDB_ENTER, ("pmap_enter: new pte 0x%08x\n", npte)); } -#endif - if (pa != pfn_to_pad(pte->pt_entry)) { + if ((pte->pt_entry & PG_V) && pa != pfn_to_pad(pte->pt_entry)) { pmap_remove(pmap, va, va + NBPG); -#ifdef PMAPDEBUG stat_count(enter_stats.mchange); -#endif } if (!(pte->pt_entry & PG_V)) { @@ -1003,15 +811,20 @@ pmap_enter(pmap, va, pa, prot, flags) pte->pt_entry = npte; if (pmap->pm_tlbgen == tlbpid_gen) { tlb_update(va | (pmap->pm_tlbpid << VMTLB_PID_SHIFT), npte); - - /* - * If mapping a memory space address invalidate ICache. - */ - if (pg != NULL) { - Mips_InvalidateICachePage(va); - } } + /* + * If mapping a memory space address invalidate ICache. + */ +#if 0 + /* XXX The following test have previously been unreliable!!! */ + if (pg != NULL && + (prot & (VM_PROT_READ | VM_PROT_EXECUTE)) == + (VM_PROT_READ | VM_PROT_EXECUTE)) +#endif + if (pg != NULL) + Mips_InvalidateICachePage(va); + return 0; } @@ -1024,11 +837,8 @@ pmap_kenter_pa(va, pa, prot) pt_entry_t *pte; u_int npte; -#ifdef PMAPDEBUG - if (pmapdebug & (PDB_FOLLOW|PDB_ENTER)) { - printf("pmap_kenter_pa(%lx, %lx, %x)\n", va, pa, prot); - } -#endif + DPRINTF(PDB_FOLLOW|PDB_ENTER, + ("pmap_kenter_pa(%p, %p, 0x%x)\n", va, pa, prot)); npte = vad_to_pfn(pa) | PG_G; if (prot & VM_PROT_WRITE) @@ -1040,33 +850,11 @@ pmap_kenter_pa(va, pa, prot) tlb_update(va, npte); } -void -pmap_kenter_cache(va, pa, prot, cache) - vaddr_t va; - paddr_t pa; - vm_prot_t prot; - int cache; -{ - pt_entry_t *pte; - u_int npte; - -#ifdef PMAPDEBUG - if (pmapdebug & (PDB_FOLLOW|PDB_ENTER)) { - printf("pmap_kenter_cache(%lx, %lx, %x)\n", va, pa, prot); - } -#endif - - npte = vad_to_pfn(pa) | PG_G; - if (prot & VM_PROT_WRITE) { - npte |= PG_M | cache; - } else { - npte |= PG_RO | cache; - } - pte = kvtopte(va); - pte->pt_entry = npte; - tlb_update(va, npte); -} - +/* + * Remove a mapping from the kernel map table. When doing this + * the cache must be synced for the VA mapped since we mapped + * pages behind the back of the VP tracking system. + */ void pmap_kremove(va, len) vaddr_t va; @@ -1076,19 +864,14 @@ pmap_kremove(va, len) vaddr_t eva; u_int entry; -#ifdef PMAPDEBUG - if (pmapdebug & (PDB_FOLLOW|PDB_REMOVE)) { - printf("pmap_kremove(%lx, %lx)\n", va, len); - } -#endif + DPRINTF(PDB_FOLLOW|PDB_REMOVE, ("pmap_kremove(%p, %p)\n", va, len)); pte = kvtopte(va); eva = va + len; for (; va < eva; va += PAGE_SIZE, pte++) { entry = pte->pt_entry; - if (entry & PG_V) { + if (!(entry & PG_V)) continue; - } Mips_SyncDCachePage(va); pte->pt_entry = PG_NV | PG_G; tlb_flush_addr(va); @@ -1118,19 +901,14 @@ pmap_extract(pmap, va, pa) boolean_t rv = TRUE; pt_entry_t *pte; -#ifdef PMAPDEBUG - if (pmapdebug & PDB_FOLLOW) { - printf("pmap_extract(%x, %x) -> ", pmap, va); - } -#endif - if (pmap == pmap_kernel()) { - if (va >= (long)KSEG0_BASE && va < (long)(KSEG0_BASE + KSEG_SIZE)) { + if (va >= (long)KSEG0_BASE && + va < (long)(KSEG0_BASE + KSEG_SIZE)) { *pa = (long)KSEG0_TO_PHYS(va); } else { #ifdef DIAGNOSTIC if (va < VM_MIN_KERNEL_ADDRESS || va >= virtual_end) { - panic("pmap_extract(%x, %x)", pmap, va); + panic("pmap_extract(%p, %p)", pmap, va); } #endif pte = kvtopte(va); @@ -1150,11 +928,7 @@ pmap_extract(pmap, va, pa) if (rv != FALSE) *pa |= va & PGOFSET; -#ifdef PMAPDEBUG - if (pmapdebug & PDB_FOLLOW) { - printf("pmap_extract: pa %x\n", *pa); - } -#endif + DPRINTF(PDB_FOLLOW, ("pmap_extract(%p, %p)=%p(%d)", pmap, va, *pa, rv)); return (rv); } @@ -1168,7 +942,11 @@ pmap_prefer(foff, vap) paddr_t foff; vaddr_t *vap; { +#if 1 *vap += (foff - *vap) & (CpuCacheAliasMask | PAGE_MASK); +#else + *vap += (*vap ^ foff) & CpuCacheAliasMask; +#endif } /* @@ -1187,12 +965,8 @@ pmap_copy(dst_pmap, src_pmap, dst_addr, len, src_addr) vaddr_t src_addr; { -#ifdef PMAPDEBUG - if (pmapdebug & PDB_FOLLOW) { - printf("pmap_copy(%x, %x, %x, %x, %x)\n", - dst_pmap, src_pmap, dst_addr, len, src_addr); - } -#endif + DPRINTF(PDB_FOLLOW, ("pmap_copy(%p, %p, %p, %p, %p)\n", + dst_pmap, src_pmap, dst_addr, len, src_addr)); } /* @@ -1205,11 +979,7 @@ pmap_zero_page(struct vm_page *pg) vaddr_t va; pv_entry_t pv; -#ifdef PMAPDEBUG - if (pmapdebug & PDB_FOLLOW) { - printf("pmap_zero_page(%x)\n", phys); - } -#endif + DPRINTF(PDB_FOLLOW, ("pmap_zero_page(%p)\n", phys)); va = (vaddr_t)PHYS_TO_KSEG0(phys); pv = pg_to_pvh(pg); @@ -1231,88 +1001,102 @@ pmap_zero_page(struct vm_page *pg) void pmap_copy_page(struct vm_page *srcpg, struct vm_page *dstpg) { - paddr_t src = VM_PAGE_TO_PHYS(srcpg); - paddr_t dst = VM_PAGE_TO_PHYS(dstpg); -#if 0 - int *s, *d; -/* if (CpuCacheAliasMask == 0) { XXX */ - s = (int *)PHYS_TO_KSEG0(src); - d = (int *)PHYS_TO_KSEG0(dst); - - memcpy(d, s, PAGE_SIZE); - -#else - int *s, *d, *end; + paddr_t src, dst; + vaddr_t s, d; int df = 1; int sf = 1; - int tmp0, tmp1, tmp2, tmp3; pv_entry_t pv; -#ifdef PMAPDEBUG - if (pmapdebug & PDB_FOLLOW) { - printf("pmap_copy_page(%x, %x)\n", src, dst); - } -#endif - s = (int *)(vaddr_t)PHYS_TO_KSEG0(src); - d = (int *)(vaddr_t)PHYS_TO_KSEG0(dst); + src = VM_PAGE_TO_PHYS(srcpg); + dst = VM_PAGE_TO_PHYS(dstpg); + s = (vaddr_t)PHYS_TO_KSEG0(src); + d = (vaddr_t)PHYS_TO_KSEG0(dst); + + DPRINTF(PDB_FOLLOW, ("pmap_copy_page(%p, %p)\n", src, dst)); pv = pg_to_pvh(srcpg); if ((pv->pv_flags & PV_CACHED) && - (sf = ((pv->pv_va ^ (long)s) & CpuCacheAliasMask) != 0)) { + (sf = ((pv->pv_va ^ (long)s) & CpuCacheAliasMask) != 0)) { Mips_SyncDCachePage(pv->pv_va); } pv = pg_to_pvh(dstpg); if ((pv->pv_flags & PV_CACHED) && - (df = ((pv->pv_va ^ (long)d) & CpuCacheAliasMask) != 0)) { + (df = ((pv->pv_va ^ (long)d) & CpuCacheAliasMask) != 0)) { Mips_SyncDCachePage(pv->pv_va); } - end = s + PAGE_SIZE / sizeof(int); - do { - tmp0 = s[0]; tmp1 = s[1]; tmp2 = s[2]; tmp3 = s[3]; - d[0] = tmp0; d[1] = tmp1; d[2] = tmp2; d[3] = tmp3; - s += 4; - d += 4; - } while (s != end); + memcpy((void *)d, (void *)s, PAGE_SIZE); if (sf) { - Mips_HitSyncDCache((vaddr_t)PHYS_TO_KSEG0(src), PAGE_SIZE); + Mips_HitSyncDCache(s, PAGE_SIZE); } #if 0 /* XXX TODO: Why can't we trust the following? */ if (df || (pv->pv_pmap == NULL) || (pv->pv_flags & PV_EXEC)) { - Mips_HitSyncDCachePage(dst); + Mips_SyncDCachePage(d); } #else - Mips_HitSyncDCache((vaddr_t)PHYS_TO_KSEG0(dst), PAGE_SIZE); -#endif + Mips_SyncDCachePage(d); #endif } /* - * Clear the modify bits on the specified physical page. + * Clear the modify bits on the specified physical page. + * Also sync the cache so it reflects the new clean state of the page. */ boolean_t -pmap_clear_modify(pg) - struct vm_page *pg; +pmap_clear_modify(struct vm_page *pg) { -#ifdef PMAPDEBUG - paddr_t pa = VM_PAGE_TO_PHYS(pg); -#endif + pv_entry_t pv; + pt_entry_t *pte; + unsigned entry; boolean_t rv = FALSE; -#ifdef PMAPDEBUG - if (pmapdebug & PDB_FOLLOW) { - printf("pmap_clear_modify(%x)\n", pa); + DPRINTF(PDB_FOLLOW, ("pmap_clear_modify(%p)\n", VM_PAGE_TO_PHYS(pg))); + + pv = pg_to_pvh(pg); + if (pv->pv_flags & PV_ATTR_MOD) { + pv->pv_flags &= ~PV_ATTR_MOD; + rv = TRUE; } -#endif - return(rv); + Mips_SyncDCachePage(pv->pv_va); + + for (; pv != NULL; pv = pv->pv_next) { + if (pv->pv_pmap == pmap_kernel()) { + pte = kvtopte(pv->pv_va); + entry = pte->pt_entry; + if ((entry & PG_V) != 0 && (entry & PG_M) != 0) { + rv = TRUE; + entry &= ~PG_M; + pte->pt_entry = entry; + tlb_update(pv->pv_va, entry); + } + } else if (pv->pv_pmap != NULL) { + if ((pte = pmap_segmap(pv->pv_pmap, pv->pv_va)) == NULL) + continue; + pte += uvtopte(pv->pv_va); + entry = pte->pt_entry; + if ((entry & PG_V) != 0 && (entry & PG_M) != 0) { + rv = TRUE; + entry &= ~PG_M; + pte->pt_entry = entry; + if (pv->pv_pmap->pm_tlbgen == tlbpid_gen) + tlb_update(pv->pv_va | (pv->pv_pmap->pm_tlbpid << + VMTLB_PID_SHIFT), entry); + } + } + } + + return rv; } void pmap_set_modify(pg) struct vm_page *pg; { - pg->flags &= ~PG_CLEAN; + pv_entry_t pv; + + pv = pg_to_pvh(pg); + pv->pv_flags |= PV_ATTR_MOD | PV_ATTR_REF; } /* @@ -1324,16 +1108,15 @@ boolean_t pmap_clear_reference(pg) struct vm_page *pg; { -#ifdef PMAPDEBUG - paddr_t pa = VM_PAGE_TO_PHYS(pg); -#endif + pv_entry_t pv; + boolean_t rv; -#ifdef PMAPDEBUG - if (pmapdebug & PDB_FOLLOW) { - printf("pmap_clear_reference(%x)\n", pa); - } -#endif - return(FALSE); + DPRINTF(PDB_FOLLOW, ("pmap_clear_reference(%p)\n", VM_PAGE_TO_PHYS(pg))); + + pv = pg_to_pvh(pg); + rv = (pv->pv_flags & PV_ATTR_REF) != 0; + pv->pv_flags &= ~PV_ATTR_REF; + return rv; } /* @@ -1346,7 +1129,10 @@ boolean_t pmap_is_referenced(pg) struct vm_page *pg; { - return (FALSE); + pv_entry_t pv; + + pv = pg_to_pvh(pg); + return (pv->pv_flags & PV_ATTR_REF) != 0; } /* @@ -1359,25 +1145,97 @@ boolean_t pmap_is_modified(pg) struct vm_page *pg; { - return (FALSE); + pv_entry_t pv; + + pv = pg_to_pvh(pg); + return (pv->pv_flags & PV_ATTR_MOD) != 0; } -paddr_t -pmap_phys_address(ppn) - int ppn; +/* + * Miscellaneous support routines not part of the pmap API + */ + +/* + * Return RO protection of page. + */ +int +pmap_is_page_ro(pmap_t pmap, vaddr_t va, int entry) { + return (entry & PG_RO); +} -#ifdef PMAPDEBUG - if (pmapdebug & PDB_FOLLOW) { - printf("pmap_phys_address(%x)\n", ppn); + +/* + * Walk the PV tree for a physical page and change all its + * mappings to cached or uncached. + */ +void +pmap_page_cache(vm_page_t pg, int mode) +{ + pv_entry_t pv; + pt_entry_t *pte; + u_int entry; + u_int newmode; + + DPRINTF(PDB_FOLLOW|PDB_ENTER, ("pmap_page_uncache(%p)\n", pg)); + + newmode = mode & PV_UNCACHED ? PG_UNCACHED : PG_CACHED; + pv = pg_to_pvh(pg); + pv->pv_flags = (pv->pv_flags & ~(PV_CACHED|PV_UNCACHED)) | mode; + + for (; pv != NULL; pv = pv->pv_next) { + if (pv->pv_pmap == pmap_kernel()) { + pte = kvtopte(pv->pv_va); + entry = pte->pt_entry; + if (entry & PG_V) { + entry = (entry & ~PG_CACHEMODE) | newmode; + pte->pt_entry = entry; + tlb_update(pv->pv_va, entry); + } + } else { + if ((pte = pmap_segmap(pv->pv_pmap, pv->pv_va))) { + pte += uvtopte(pv->pv_va); + entry = pte->pt_entry; + if (entry & PG_V) { + entry = (entry & ~PG_CACHEMODE) | newmode; + pte->pt_entry = entry; + if (pv->pv_pmap->pm_tlbgen == tlbpid_gen) + tlb_update(pv->pv_va | (pv->pv_pmap->pm_tlbpid << + VMTLB_PID_SHIFT), entry); + } + } + } } -#endif - return (ptoa(ppn)); } /* - * Miscellaneous support routines + * Use this function to allocate pages for the mapping tables. + * Mapping tables are walked by the TLB miss code and are mapped in + * KSEG0 to avoid additional page faults when servicing a TLB miss. */ +int +pmap_page_alloc(vaddr_t *ret) +{ + vm_page_t pg; + pv_entry_t pv; + vaddr_t va; + + pg = uvm_pagealloc(NULL, 0, NULL, UVM_PGA_USERESERVE | UVM_PGA_ZERO); + if (pg == NULL) + return ENOMEM; + + pv = pg_to_pvh(pg); + va = PHYS_TO_KSEG0(VM_PAGE_TO_PHYS(pg)); + if ((pv->pv_flags & PV_CACHED) && + ((pv->pv_va ^ va) & CpuCacheAliasMask) != 0) { + Mips_SyncDCachePage(pv->pv_va); + } + pv->pv_va = va; + pv->pv_flags = PV_CACHED; + + *ret = va; + return 0; +} /* * Allocate a hardware PID and return it. @@ -1412,44 +1270,125 @@ pmap_alloc_tlbpid(p) id = pmap->pm_tlbpid; } + if (curproc) { + DPRINTF(PDB_FOLLOW|PDB_TLBPID, + ("pmap_alloc_tlbpid: curproc %d '%s' ", + curproc->p_pid, curproc->p_comm)); + } else { + DPRINTF(PDB_FOLLOW|PDB_TLBPID, + ("pmap_alloc_tlbpid: curproc <none> ")); + } + DPRINTF(PDB_FOLLOW|PDB_TLBPID, ("segtab %p tlbpid %d pid %d '%s'\n", + pmap->pm_segtab, id, p->p_pid, p->p_comm)); + + return (id); +} + +/* + * Enter the pmap and virtual address into the physical to virtual map table. + */ +int +pmap_enter_pv(pmap_t pmap, vaddr_t va, vm_page_t pg, u_int *npte) +{ + pv_entry_t pv, npv; + + pv = pg_to_pvh(pg); + + if (pv->pv_pmap == NULL) { + /* + * No entries yet, use header as the first entry + */ + + DPRINTF(PDB_PVENTRY, + ("pmap_enter: first pv: pmap %p va %p pa %p\n", + pmap, va, VM_PAGE_TO_PHYS(pg))); + + stat_count(enter_stats.firstpv); + + Mips_SyncDCachePage(pv->pv_va); + + pv->pv_va = va; + pv->pv_flags = PV_CACHED; + pv->pv_pmap = pmap; + pv->pv_next = NULL; + } else { + if (pv->pv_flags & PV_UNCACHED) { + /* + * If page is mapped uncached it's either because + * an uncached mapping was requested of we have a + * VAC situation. Map this page uncached as well. + */ + *npte = (*npte & ~PG_CACHEMODE) | PG_UNCACHED; + } else if (CpuCacheAliasMask != 0) { + /* + * We have a VAC possibility. Check if virtual + * address of current mappings are compatible + * with this new mapping. Only need to check first + * since all others have been checked compatible + * when added. If they are incompatible, remove + * all mappings, flush the cache and set page + * to be mapped uncached. + */ + if (((pv->pv_va ^ va) & CpuCacheAliasMask) != 0) { #ifdef PMAPDEBUG - if (pmapdebug & (PDB_FOLLOW|PDB_TLBPID)) { - if (curproc) { - printf("pmap_alloc_tlbpid: curproc %d '%s' ", - curproc->p_pid, curproc->p_comm); - } else { - printf("pmap_alloc_tlbpid: curproc <none> "); + printf("pmap_enter: VAC for pa %p, %p != %p\n", + VM_PAGE_TO_PHYS(pg), npv->pv_va, va); +#endif + pmap_page_cache(pg, PV_UNCACHED); + Mips_SyncCache(); + *npte = (*npte & ~PG_CACHEMODE) | PG_UNCACHED; + } + } + + /* + * There is at least one other VA mapping this page. + * Place this entry after the header. + * + * Note: the entry may already be in the table if + * we are only changing the protection bits. + */ + for (npv = pv; npv; npv = npv->pv_next) { + if (pmap == npv->pv_pmap && va == npv->pv_va) { + return 0; + } } - printf("segtab %x tlbpid %d pid %d '%s'\n", - pmap->pm_segtab, id, p->p_pid, p->p_comm); + + DPRINTF(PDB_PVENTRY, + ("pmap_enter: new pv: pmap %x va %x pg %p\n", + pmap, va, VM_PAGE_TO_PHYS(pg))); + + /* can this cause us to recurse forever? */ + npv = pmap_pv_alloc(); + if (npv == NULL) + return ENOMEM; + npv->pv_va = va; + npv->pv_pmap = pmap; + npv->pv_next = pv->pv_next; + npv->pv_flags = pv->pv_flags; + pv->pv_next = npv; + + if (!npv->pv_next) + stat_count(enter_stats.secondpv); } -#endif - return (id); + return 0; } /* - * Remove a physical to virtual address translation. + * Remove a physical to virtual address translation from the PV table. * Returns TRUE if it was the last mapping and cached, else FALSE. */ void -pmap_remove_pv(pmap, va, pa) - pmap_t pmap; - vaddr_t va; - paddr_t pa; +pmap_remove_pv(pmap_t pmap, vaddr_t va, paddr_t pa) { pv_entry_t pv, npv; vm_page_t pg; -#ifdef PMAPDEBUG - if (pmapdebug & (PDB_FOLLOW|PDB_PVENTRY)) { - printf("pmap_remove_pv(%x, %x, %x)\n", pmap, va, pa); - } -#endif + DPRINTF(PDB_FOLLOW|PDB_PVENTRY, + ("pmap_remove_pv(%p, %p, %p)\n", pmap, va, pa)); /* - * Remove page from the PV table (raise IPL since we - * may be called at interrupt time). + * Remove page from the PV table */ pg = PHYS_TO_VM_PAGE(pa); if (pg == NULL) @@ -1465,29 +1404,29 @@ pmap_remove_pv(pmap, va, pa) if (pmap == pv->pv_pmap && va == pv->pv_va) { npv = pv->pv_next; if (npv) { + npv->pv_flags |= pv->pv_flags & PV_PRESERVE; *pv = *npv; pmap_pv_free(npv); } else { pv->pv_pmap = NULL; + Mips_SyncDCachePage(pv->pv_va); } -#ifdef PMAPDEBUG stat_count(remove_stats.pvfirst); -#endif - } - else { + } else { for (npv = pv->pv_next; npv; pv = npv, npv = npv->pv_next) { -#ifdef PMAPDEBUG stat_count(remove_stats.pvsearch); -#endif if (pmap == npv->pv_pmap && va == npv->pv_va) - goto fnd; + break; } + if (npv != NULL) { + pv->pv_next = npv->pv_next; + pmap_pv_free(npv); + } else { #ifdef DIAGNOSTIC - panic("pmap_remove_pv(%x, %x, %x) not found\n", pmap, va, pa); + panic("pmap_remove_pv(%x, %x, %x) not found", + pmap, va, pa); #endif - fnd: - pv->pv_next = npv->pv_next; - pmap_pv_free(npv); + } } } @@ -1519,9 +1458,15 @@ bus_mem_add_mapping(bus_addr_t bpa, bus_size_t size, int cacheable, printf("map bus %x size %x to %x vbase %x\n", bpa, size, *bshp, spa); #endif for (; len > 0; len -= NBPG) { - pmap_kenter_cache(vaddr, spa, - VM_PROT_READ | VM_PROT_WRITE, - cacheable ? PG_IOPAGE : PG_IOPAGE); /* XXX */ + pt_entry_t *pte; + u_int npte; + + npte = vad_to_pfn(spa) | PG_G; + npte |= PG_V | PG_M | PG_IOPAGE; + pte = kvtopte(vaddr); + pte->pt_entry = npte; + tlb_update(vaddr, npte); + spa += NBPG; vaddr += NBPG; } diff --git a/sys/arch/mips64/mips64/trap.c b/sys/arch/mips64/mips64/trap.c index 870c9d6e981..b7def1b6d7d 100644 --- a/sys/arch/mips64/mips64/trap.c +++ b/sys/arch/mips64/mips64/trap.c @@ -1,4 +1,4 @@ -/* $OpenBSD: trap.c,v 1.11 2004/09/22 14:39:44 miod Exp $ */ +/* $OpenBSD: trap.c,v 1.12 2004/09/23 08:42:38 pefo Exp $ */ /* tracked to 1.23 */ /* @@ -216,8 +216,7 @@ trap(trapframe) } entry |= PG_M; pte->pt_entry = entry; - trapframe->badvaddr &= ~PGOFSET; - tlb_update(trapframe->badvaddr, entry); + tlb_update(trapframe->badvaddr & ~PGOFSET, entry); pa = pfn_to_pad(entry); pg = PHYS_TO_VM_PAGE(pa); if (pg == NULL) @@ -237,7 +236,7 @@ trap(trapframe) if (!(pte = pmap_segmap(pmap, trapframe->badvaddr))) panic("trap: utlbmod: invalid segmap"); - pte += (trapframe->badvaddr >> PGSHIFT) & (NPTEPG - 1); + pte += uvtopte(trapframe->badvaddr); entry = pte->pt_entry; #ifdef DIAGNOSTIC if (!(entry & PG_V) || (entry & PG_M)) @@ -251,9 +250,8 @@ trap(trapframe) } entry |= PG_M; pte->pt_entry = entry; - trapframe->badvaddr = (trapframe->badvaddr & ~PGOFSET) | - (pmap->pm_tlbpid << VMTLB_PID_SHIFT); - tlb_update(trapframe->badvaddr, entry); + tlb_update((trapframe->badvaddr & ~PGOFSET) | + (pmap->pm_tlbpid << VMTLB_PID_SHIFT), entry); pa = pfn_to_pad(entry); pg = PHYS_TO_VM_PAGE(pa); if (pg == NULL) @@ -587,6 +585,7 @@ printf("SIG-BUSB @%p pc %p, ra %p\n", trapframe->badvaddr, trapframe->pc, trapfr u_int32_t instr; struct uio uio; struct iovec iov; + struct trap_frame *locr0 = p->p_md.md_regs; /* compute address of break instruction */ va = (caddr_t)trapframe->pc; @@ -594,40 +593,70 @@ printf("SIG-BUSB @%p pc %p, ra %p\n", trapframe->badvaddr, trapframe->pc, trapfr va += 4; /* read break instruction */ - copyin(&instr, va, sizeof(int32_t)); + copyin(va, &instr, sizeof(int32_t)); + #if 0 printf("trap: %s (%d) breakpoint %x at %x: (adr %x ins %x)\n", p->p_comm, p->p_pid, instr, trapframe->pc, p->p_md.md_ss_addr, p->p_md.md_ss_instr); /* XXX */ #endif - if (p->p_md.md_ss_addr != (long)va || instr != BREAK_SSTEP) { + + switch ((instr & BREAK_VAL_MASK) >> BREAK_VAL_SHIFT) { + case 6: /* gcc range error */ + i = SIGFPE; + typ = FPE_FLTSUB; + /* skip instruction */ + if ((int)trapframe->cause & CR_BR_DELAY) + locr0->pc = MipsEmulateBranch(locr0, + trapframe->pc, 0, 0); + else + locr0->pc += 4; + break; + case 7: /* gcc divide by zero */ + i = SIGFPE; + typ = FPE_FLTDIV; /* XXX FPE_INTDIV ? */ + /* skip instruction */ + if ((int)trapframe->cause & CR_BR_DELAY) + locr0->pc = MipsEmulateBranch(locr0, + trapframe->pc, 0, 0); + else + locr0->pc += 4; + break; + case BREAK_SSTEP_VAL: + if (p->p_md.md_ss_addr == (long)va) { + /* + * Restore original instruction and clear BP + */ + iov.iov_base = (caddr_t)&p->p_md.md_ss_instr; + iov.iov_len = sizeof(int); + uio.uio_iov = &iov; + uio.uio_iovcnt = 1; + uio.uio_offset = (off_t)(long)va; + uio.uio_resid = sizeof(int); + uio.uio_segflg = UIO_SYSSPACE; + uio.uio_rw = UIO_WRITE; + uio.uio_procp = curproc; + i = procfs_domem(p, p, NULL, &uio); + Mips_SyncCache(); + + if (i < 0) + printf("Warning: can't restore instruction at %x: %x\n", + p->p_md.md_ss_addr, + p->p_md.md_ss_instr); + + p->p_md.md_ss_addr = 0; + typ = TRAP_BRKPT; + } else { + typ = TRAP_TRACE; + } i = SIGTRAP; + break; + default: typ = TRAP_TRACE; + i = SIGTRAP; break; } - /* - * Restore original instruction and clear BP - */ - iov.iov_base = (caddr_t)&p->p_md.md_ss_instr; - iov.iov_len = sizeof(int); - uio.uio_iov = &iov; - uio.uio_iovcnt = 1; - uio.uio_offset = (off_t)(long)va; - uio.uio_resid = sizeof(int); - uio.uio_segflg = UIO_SYSSPACE; - uio.uio_rw = UIO_WRITE; - uio.uio_procp = curproc; - i = procfs_domem(p, p, NULL, &uio); - Mips_SyncCache(); - - if (i < 0) - printf("Warning: can't restore instruction at %x: %x\n", - p->p_md.md_ss_addr, p->p_md.md_ss_instr); - - p->p_md.md_ss_addr = 0; - i = SIGTRAP; - typ = TRAP_BRKPT; break; } @@ -660,12 +689,11 @@ printf("SIG-BUSB @%p pc %p, ra %p\n", trapframe->badvaddr, trapframe->pc, trapfr if ((int)trapframe->cause & CR_BR_DELAY) va += 4; /* read break instruction */ - copyin(&instr, va, sizeof(int32_t)); + copyin(va, &instr, sizeof(int32_t)); if ((int)trapframe->cause & CR_BR_DELAY) { locr0->pc = MipsEmulateBranch(locr0, trapframe->pc, 0, 0); - } - else { + } else { locr0->pc += 4; } if (instr == 0x040c0000) { /* Performance cntr trap */ |