/* $Id: IEMAllCImplStrInstr.cpp.h 49271 2013-10-24 10:50:57Z vboxsync $ */ /** @file * IEM - String Instruction Implementation Code Template. */ /* * Copyright (C) 2011-2012 Oracle Corporation * * This file is part of VirtualBox Open Source Edition (OSE), as * available from http://www.virtualbox.org. This file is free software; * you can redistribute it and/or modify it under the terms of the GNU * General Public License (GPL) as published by the Free Software * Foundation, in version 2 as it comes in the "COPYING" file of the * VirtualBox OSE distribution. VirtualBox OSE is distributed in the * hope that it will be useful, but WITHOUT ANY WARRANTY of any kind. */ /******************************************************************************* * Defined Constants And Macros * *******************************************************************************/ #if OP_SIZE == 8 # define OP_rAX al #elif OP_SIZE == 16 # define OP_rAX ax #elif OP_SIZE == 32 # define OP_rAX eax #elif OP_SIZE == 64 # define OP_rAX rax #else # error "Bad OP_SIZE." #endif #define OP_TYPE RT_CONCAT3(uint,OP_SIZE,_t) #if ADDR_SIZE == 16 # define ADDR_rDI di # define ADDR_rSI si # define ADDR_rCX cx # define ADDR2_TYPE uint32_t #elif ADDR_SIZE == 32 # define ADDR_rDI edi # define ADDR_rSI esi # define ADDR_rCX ecx # define ADDR2_TYPE uint32_t #elif ADDR_SIZE == 64 # define ADDR_rDI rdi # define ADDR_rSI rsi # define ADDR_rCX rcx # define ADDR2_TYPE uint64_t # define IS_64_BIT_CODE(a_pIemCpu) (true) #else # error "Bad ADDR_SIZE." #endif #define ADDR_TYPE RT_CONCAT3(uint,ADDR_SIZE,_t) #if ADDR_SIZE == 64 || OP_SIZE == 64 # define IS_64_BIT_CODE(a_pIemCpu) (true) #elif ADDR_SIZE == 32 # define IS_64_BIT_CODE(a_pIemCpu) ((a_pIemCpu)->enmCpuMode == IEMMODE_64BIT) #else # define IS_64_BIT_CODE(a_pIemCpu) (false) #endif /** * Implements 'REPE CMPS'. */ IEM_CIMPL_DEF_1(RT_CONCAT4(iemCImpl_repe_cmps_op,OP_SIZE,_addr,ADDR_SIZE), uint8_t, iEffSeg) { PCPUMCTX pCtx = pIemCpu->CTX_SUFF(pCtx); /* * Setup. */ ADDR_TYPE uCounterReg = pCtx->ADDR_rCX; if (uCounterReg == 0) { iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } PCCPUMSELREGHID pSrc1Hid = iemSRegGetHid(pIemCpu, iEffSeg); uint64_t uSrc1Base; VBOXSTRICTRC rcStrict = iemMemSegCheckReadAccessEx(pIemCpu, pSrc1Hid, iEffSeg, &uSrc1Base); if (rcStrict != VINF_SUCCESS) return rcStrict; uint64_t uSrc2Base; rcStrict = iemMemSegCheckReadAccessEx(pIemCpu, &pCtx->es, X86_SREG_ES, &uSrc2Base); if (rcStrict != VINF_SUCCESS) return rcStrict; int8_t const cbIncr = pCtx->eflags.Bits.u1DF ? -(OP_SIZE / 8) : (OP_SIZE / 8); ADDR_TYPE uSrc1AddrReg = pCtx->ADDR_rSI; ADDR_TYPE uSrc2AddrReg = pCtx->ADDR_rDI; uint32_t uEFlags = pCtx->eflags.u; /* * The loop. */ do { /* * Do segmentation and virtual page stuff. */ ADDR2_TYPE uVirtSrc1Addr = uSrc1AddrReg + (ADDR2_TYPE)uSrc1Base; ADDR2_TYPE uVirtSrc2Addr = uSrc2AddrReg + (ADDR2_TYPE)uSrc2Base; uint32_t cLeftSrc1Page = (PAGE_SIZE - (uVirtSrc1Addr & PAGE_OFFSET_MASK)) / (OP_SIZE / 8); if (cLeftSrc1Page > uCounterReg) cLeftSrc1Page = uCounterReg; uint32_t cLeftSrc2Page = (PAGE_SIZE - (uVirtSrc2Addr & PAGE_OFFSET_MASK)) / (OP_SIZE / 8); uint32_t cLeftPage = RT_MIN(cLeftSrc1Page, cLeftSrc2Page); if ( cLeftPage > 0 /* can be null if unaligned, do one fallback round. */ && cbIncr > 0 /** @todo Optimize reverse direction string ops. */ && ( IS_64_BIT_CODE(pIemCpu) || ( uSrc1AddrReg < pSrc1Hid->u32Limit && uSrc1AddrReg + (cLeftPage * (OP_SIZE / 8)) <= pSrc1Hid->u32Limit && uSrc2AddrReg < pCtx->es.u32Limit && uSrc2AddrReg + (cLeftPage * (OP_SIZE / 8)) <= pCtx->es.u32Limit) ) ) { RTGCPHYS GCPhysSrc1Mem; rcStrict = iemMemPageTranslateAndCheckAccess(pIemCpu, uVirtSrc1Addr, IEM_ACCESS_DATA_R, &GCPhysSrc1Mem); if (rcStrict != VINF_SUCCESS) return rcStrict; RTGCPHYS GCPhysSrc2Mem; rcStrict = iemMemPageTranslateAndCheckAccess(pIemCpu, uVirtSrc2Addr, IEM_ACCESS_DATA_R, &GCPhysSrc2Mem); if (rcStrict != VINF_SUCCESS) return rcStrict; /* * If we can map the page without trouble, do a block processing * until the end of the current page. */ PGMPAGEMAPLOCK PgLockSrc2Mem; OP_TYPE const *puSrc2Mem; rcStrict = iemMemPageMap(pIemCpu, GCPhysSrc2Mem, IEM_ACCESS_DATA_R, (void **)&puSrc2Mem, &PgLockSrc2Mem); if (rcStrict == VINF_SUCCESS) { PGMPAGEMAPLOCK PgLockSrc1Mem; OP_TYPE const *puSrc1Mem; rcStrict = iemMemPageMap(pIemCpu, GCPhysSrc1Mem, IEM_ACCESS_DATA_R, (void **)&puSrc1Mem, &PgLockSrc1Mem); if (rcStrict == VINF_SUCCESS) { if (!memcmp(puSrc2Mem, puSrc1Mem, cLeftPage * (OP_SIZE / 8))) { /* All matches, only compare the last itme to get the right eflags. */ RT_CONCAT(iemAImpl_cmp_u,OP_SIZE)((OP_TYPE *)&puSrc1Mem[cLeftPage-1], puSrc2Mem[cLeftPage-1], &uEFlags); uSrc1AddrReg += cLeftPage * cbIncr; uSrc2AddrReg += cLeftPage * cbIncr; uCounterReg -= cLeftPage; } else { /* Some mismatch, compare each item (and keep volatile memory in mind). */ uint32_t off = 0; do { RT_CONCAT(iemAImpl_cmp_u,OP_SIZE)((OP_TYPE *)&puSrc1Mem[off], puSrc2Mem[off], &uEFlags); off++; } while ( off < cLeftPage && (uEFlags & X86_EFL_ZF)); uSrc1AddrReg += cbIncr * off; uSrc2AddrReg += cbIncr * off; uCounterReg -= off; } /* Update the registers before looping. */ pCtx->ADDR_rCX = uCounterReg; pCtx->ADDR_rSI = uSrc1AddrReg; pCtx->ADDR_rDI = uSrc2AddrReg; pCtx->eflags.u = uEFlags; iemMemPageUnmap(pIemCpu, GCPhysSrc1Mem, IEM_ACCESS_DATA_R, puSrc1Mem, &PgLockSrc1Mem); iemMemPageUnmap(pIemCpu, GCPhysSrc2Mem, IEM_ACCESS_DATA_R, puSrc2Mem, &PgLockSrc2Mem); continue; } iemMemPageUnmap(pIemCpu, GCPhysSrc2Mem, IEM_ACCESS_DATA_R, puSrc2Mem, &PgLockSrc2Mem); } } /* * Fallback - slow processing till the end of the current page. * In the cross page boundrary case we will end up here with cLeftPage * as 0, we execute one loop then. */ do { OP_TYPE uValue1; rcStrict = RT_CONCAT(iemMemFetchDataU,OP_SIZE)(pIemCpu, &uValue1, iEffSeg, uSrc1AddrReg); if (rcStrict != VINF_SUCCESS) return rcStrict; OP_TYPE uValue2; rcStrict = RT_CONCAT(iemMemFetchDataU,OP_SIZE)(pIemCpu, &uValue2, X86_SREG_ES, uSrc2AddrReg); if (rcStrict != VINF_SUCCESS) return rcStrict; RT_CONCAT(iemAImpl_cmp_u,OP_SIZE)(&uValue1, uValue2, &uEFlags); pCtx->ADDR_rSI = uSrc1AddrReg += cbIncr; pCtx->ADDR_rDI = uSrc2AddrReg += cbIncr; pCtx->ADDR_rCX = --uCounterReg; pCtx->eflags.u = uEFlags; cLeftPage--; } while ( (int32_t)cLeftPage > 0 && (uEFlags & X86_EFL_ZF)); } while ( uCounterReg != 0 && (uEFlags & X86_EFL_ZF)); /* * Done. */ iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } /** * Implements 'REPNE CMPS'. */ IEM_CIMPL_DEF_1(RT_CONCAT4(iemCImpl_repne_cmps_op,OP_SIZE,_addr,ADDR_SIZE), uint8_t, iEffSeg) { PCPUMCTX pCtx = pIemCpu->CTX_SUFF(pCtx); /* * Setup. */ ADDR_TYPE uCounterReg = pCtx->ADDR_rCX; if (uCounterReg == 0) { iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } PCCPUMSELREGHID pSrc1Hid = iemSRegGetHid(pIemCpu, iEffSeg); uint64_t uSrc1Base; VBOXSTRICTRC rcStrict = iemMemSegCheckReadAccessEx(pIemCpu, pSrc1Hid, iEffSeg, &uSrc1Base); if (rcStrict != VINF_SUCCESS) return rcStrict; uint64_t uSrc2Base; rcStrict = iemMemSegCheckReadAccessEx(pIemCpu, &pCtx->es, X86_SREG_ES, &uSrc2Base); if (rcStrict != VINF_SUCCESS) return rcStrict; int8_t const cbIncr = pCtx->eflags.Bits.u1DF ? -(OP_SIZE / 8) : (OP_SIZE / 8); ADDR_TYPE uSrc1AddrReg = pCtx->ADDR_rSI; ADDR_TYPE uSrc2AddrReg = pCtx->ADDR_rDI; uint32_t uEFlags = pCtx->eflags.u; /* * The loop. */ do { /* * Do segmentation and virtual page stuff. */ ADDR2_TYPE uVirtSrc1Addr = uSrc1AddrReg + (ADDR2_TYPE)uSrc1Base; ADDR2_TYPE uVirtSrc2Addr = uSrc2AddrReg + (ADDR2_TYPE)uSrc2Base; uint32_t cLeftSrc1Page = (PAGE_SIZE - (uVirtSrc1Addr & PAGE_OFFSET_MASK)) / (OP_SIZE / 8); if (cLeftSrc1Page > uCounterReg) cLeftSrc1Page = uCounterReg; uint32_t cLeftSrc2Page = (PAGE_SIZE - (uVirtSrc2Addr & PAGE_OFFSET_MASK)) / (OP_SIZE / 8); uint32_t cLeftPage = RT_MIN(cLeftSrc1Page, cLeftSrc2Page); if ( cLeftPage > 0 /* can be null if unaligned, do one fallback round. */ && cbIncr > 0 /** @todo Optimize reverse direction string ops. */ && ( IS_64_BIT_CODE(pIemCpu) || ( uSrc1AddrReg < pSrc1Hid->u32Limit && uSrc1AddrReg + (cLeftPage * (OP_SIZE / 8)) <= pSrc1Hid->u32Limit && uSrc2AddrReg < pCtx->es.u32Limit && uSrc2AddrReg + (cLeftPage * (OP_SIZE / 8)) <= pCtx->es.u32Limit) ) ) { RTGCPHYS GCPhysSrc1Mem; rcStrict = iemMemPageTranslateAndCheckAccess(pIemCpu, uVirtSrc1Addr, IEM_ACCESS_DATA_R, &GCPhysSrc1Mem); if (rcStrict != VINF_SUCCESS) return rcStrict; RTGCPHYS GCPhysSrc2Mem; rcStrict = iemMemPageTranslateAndCheckAccess(pIemCpu, uVirtSrc2Addr, IEM_ACCESS_DATA_R, &GCPhysSrc2Mem); if (rcStrict != VINF_SUCCESS) return rcStrict; /* * If we can map the page without trouble, do a block processing * until the end of the current page. */ OP_TYPE const *puSrc2Mem; PGMPAGEMAPLOCK PgLockSrc2Mem; rcStrict = iemMemPageMap(pIemCpu, GCPhysSrc2Mem, IEM_ACCESS_DATA_R, (void **)&puSrc2Mem, &PgLockSrc2Mem); if (rcStrict == VINF_SUCCESS) { OP_TYPE const *puSrc1Mem; PGMPAGEMAPLOCK PgLockSrc1Mem; rcStrict = iemMemPageMap(pIemCpu, GCPhysSrc1Mem, IEM_ACCESS_DATA_R, (void **)&puSrc1Mem, &PgLockSrc1Mem); if (rcStrict == VINF_SUCCESS) { if (memcmp(puSrc2Mem, puSrc1Mem, cLeftPage * (OP_SIZE / 8))) { /* All matches, only compare the last item to get the right eflags. */ RT_CONCAT(iemAImpl_cmp_u,OP_SIZE)((OP_TYPE *)&puSrc1Mem[cLeftPage-1], puSrc2Mem[cLeftPage-1], &uEFlags); uSrc1AddrReg += cLeftPage * cbIncr; uSrc2AddrReg += cLeftPage * cbIncr; uCounterReg -= cLeftPage; } else { /* Some mismatch, compare each item (and keep volatile memory in mind). */ uint32_t off = 0; do { RT_CONCAT(iemAImpl_cmp_u,OP_SIZE)((OP_TYPE *)&puSrc1Mem[off], puSrc2Mem[off], &uEFlags); off++; } while ( off < cLeftPage && !(uEFlags & X86_EFL_ZF)); uSrc1AddrReg += cbIncr * off; uSrc2AddrReg += cbIncr * off; uCounterReg -= off; } /* Update the registers before looping. */ pCtx->ADDR_rCX = uCounterReg; pCtx->ADDR_rSI = uSrc1AddrReg; pCtx->ADDR_rDI = uSrc2AddrReg; pCtx->eflags.u = uEFlags; iemMemPageUnmap(pIemCpu, GCPhysSrc1Mem, IEM_ACCESS_DATA_R, puSrc1Mem, &PgLockSrc1Mem); iemMemPageUnmap(pIemCpu, GCPhysSrc2Mem, IEM_ACCESS_DATA_R, puSrc2Mem, &PgLockSrc2Mem); continue; } iemMemPageUnmap(pIemCpu, GCPhysSrc2Mem, IEM_ACCESS_DATA_R, puSrc2Mem, &PgLockSrc2Mem); } } /* * Fallback - slow processing till the end of the current page. * In the cross page boundrary case we will end up here with cLeftPage * as 0, we execute one loop then. */ do { OP_TYPE uValue1; rcStrict = RT_CONCAT(iemMemFetchDataU,OP_SIZE)(pIemCpu, &uValue1, iEffSeg, uSrc1AddrReg); if (rcStrict != VINF_SUCCESS) return rcStrict; OP_TYPE uValue2; rcStrict = RT_CONCAT(iemMemFetchDataU,OP_SIZE)(pIemCpu, &uValue2, X86_SREG_ES, uSrc2AddrReg); if (rcStrict != VINF_SUCCESS) return rcStrict; RT_CONCAT(iemAImpl_cmp_u,OP_SIZE)(&uValue1, uValue2, &uEFlags); pCtx->ADDR_rSI = uSrc1AddrReg += cbIncr; pCtx->ADDR_rDI = uSrc2AddrReg += cbIncr; pCtx->ADDR_rCX = --uCounterReg; pCtx->eflags.u = uEFlags; cLeftPage--; } while ( (int32_t)cLeftPage > 0 && !(uEFlags & X86_EFL_ZF)); } while ( uCounterReg != 0 && !(uEFlags & X86_EFL_ZF)); /* * Done. */ iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } /** * Implements 'REPE SCAS'. */ IEM_CIMPL_DEF_0(RT_CONCAT4(iemCImpl_repe_scas_,OP_rAX,_m,ADDR_SIZE)) { PCPUMCTX pCtx = pIemCpu->CTX_SUFF(pCtx); /* * Setup. */ ADDR_TYPE uCounterReg = pCtx->ADDR_rCX; if (uCounterReg == 0) { iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } uint64_t uBaseAddr; VBOXSTRICTRC rcStrict = iemMemSegCheckReadAccessEx(pIemCpu, &pCtx->es, X86_SREG_ES, &uBaseAddr); if (rcStrict != VINF_SUCCESS) return rcStrict; int8_t const cbIncr = pCtx->eflags.Bits.u1DF ? -(OP_SIZE / 8) : (OP_SIZE / 8); OP_TYPE const uValueReg = pCtx->OP_rAX; ADDR_TYPE uAddrReg = pCtx->ADDR_rDI; uint32_t uEFlags = pCtx->eflags.u; /* * The loop. */ do { /* * Do segmentation and virtual page stuff. */ ADDR2_TYPE uVirtAddr = uAddrReg + (ADDR2_TYPE)uBaseAddr; uint32_t cLeftPage = (PAGE_SIZE - (uVirtAddr & PAGE_OFFSET_MASK)) / (OP_SIZE / 8); if (cLeftPage > uCounterReg) cLeftPage = uCounterReg; if ( cLeftPage > 0 /* can be null if unaligned, do one fallback round. */ && cbIncr > 0 /** @todo Implement reverse direction string ops. */ && ( IS_64_BIT_CODE(pIemCpu) || ( uAddrReg < pCtx->es.u32Limit && uAddrReg + (cLeftPage * (OP_SIZE / 8)) <= pCtx->es.u32Limit) ) ) { RTGCPHYS GCPhysMem; rcStrict = iemMemPageTranslateAndCheckAccess(pIemCpu, uVirtAddr, IEM_ACCESS_DATA_R, &GCPhysMem); if (rcStrict != VINF_SUCCESS) return rcStrict; /* * If we can map the page without trouble, do a block processing * until the end of the current page. */ PGMPAGEMAPLOCK PgLockMem; OP_TYPE const *puMem; rcStrict = iemMemPageMap(pIemCpu, GCPhysMem, IEM_ACCESS_DATA_R, (void **)&puMem, &PgLockMem); if (rcStrict == VINF_SUCCESS) { /* Search till we find a mismatching item. */ OP_TYPE uTmpValue; bool fQuit; uint32_t i = 0; do { uTmpValue = puMem[i++]; fQuit = uTmpValue != uValueReg; } while (i < cLeftPage && !fQuit); /* Update the regs. */ RT_CONCAT(iemAImpl_cmp_u,OP_SIZE)((OP_TYPE *)&uValueReg, uTmpValue, &uEFlags); pCtx->ADDR_rCX = uCounterReg -= i; pCtx->ADDR_rDI = uAddrReg += i * cbIncr; pCtx->eflags.u = uEFlags; Assert(!(uEFlags & X86_EFL_ZF) == fQuit); iemMemPageUnmap(pIemCpu, GCPhysMem, IEM_ACCESS_DATA_R, puMem, &PgLockMem); if (fQuit) break; /* If unaligned, we drop thru and do the page crossing access below. Otherwise, do the next page. */ if (!(uVirtAddr & (OP_SIZE - 1))) continue; if (uCounterReg == 0) break; cLeftPage = 0; } } /* * Fallback - slow processing till the end of the current page. * In the cross page boundrary case we will end up here with cLeftPage * as 0, we execute one loop then. */ do { OP_TYPE uTmpValue; rcStrict = RT_CONCAT(iemMemFetchDataU,OP_SIZE)(pIemCpu, &uTmpValue, X86_SREG_ES, uAddrReg); if (rcStrict != VINF_SUCCESS) return rcStrict; RT_CONCAT(iemAImpl_cmp_u,OP_SIZE)((OP_TYPE *)&uValueReg, uTmpValue, &uEFlags); pCtx->ADDR_rDI = uAddrReg += cbIncr; pCtx->ADDR_rCX = --uCounterReg; pCtx->eflags.u = uEFlags; cLeftPage--; } while ( (int32_t)cLeftPage > 0 && (uEFlags & X86_EFL_ZF)); } while ( uCounterReg != 0 && (uEFlags & X86_EFL_ZF)); /* * Done. */ iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } /** * Implements 'REPNE SCAS'. */ IEM_CIMPL_DEF_0(RT_CONCAT4(iemCImpl_repne_scas_,OP_rAX,_m,ADDR_SIZE)) { PCPUMCTX pCtx = pIemCpu->CTX_SUFF(pCtx); /* * Setup. */ ADDR_TYPE uCounterReg = pCtx->ADDR_rCX; if (uCounterReg == 0) { iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } uint64_t uBaseAddr; VBOXSTRICTRC rcStrict = iemMemSegCheckReadAccessEx(pIemCpu, &pCtx->es, X86_SREG_ES, &uBaseAddr); if (rcStrict != VINF_SUCCESS) return rcStrict; int8_t const cbIncr = pCtx->eflags.Bits.u1DF ? -(OP_SIZE / 8) : (OP_SIZE / 8); OP_TYPE const uValueReg = pCtx->OP_rAX; ADDR_TYPE uAddrReg = pCtx->ADDR_rDI; uint32_t uEFlags = pCtx->eflags.u; /* * The loop. */ do { /* * Do segmentation and virtual page stuff. */ ADDR2_TYPE uVirtAddr = uAddrReg + (ADDR2_TYPE)uBaseAddr; uint32_t cLeftPage = (PAGE_SIZE - (uVirtAddr & PAGE_OFFSET_MASK)) / (OP_SIZE / 8); if (cLeftPage > uCounterReg) cLeftPage = uCounterReg; if ( cLeftPage > 0 /* can be null if unaligned, do one fallback round. */ && cbIncr > 0 /** @todo Implement reverse direction string ops. */ && ( IS_64_BIT_CODE(pIemCpu) || ( uAddrReg < pCtx->es.u32Limit && uAddrReg + (cLeftPage * (OP_SIZE / 8)) <= pCtx->es.u32Limit) ) ) { RTGCPHYS GCPhysMem; rcStrict = iemMemPageTranslateAndCheckAccess(pIemCpu, uVirtAddr, IEM_ACCESS_DATA_R, &GCPhysMem); if (rcStrict != VINF_SUCCESS) return rcStrict; /* * If we can map the page without trouble, do a block processing * until the end of the current page. */ PGMPAGEMAPLOCK PgLockMem; OP_TYPE const *puMem; rcStrict = iemMemPageMap(pIemCpu, GCPhysMem, IEM_ACCESS_DATA_R, (void **)&puMem, &PgLockMem); if (rcStrict == VINF_SUCCESS) { /* Search till we find a mismatching item. */ OP_TYPE uTmpValue; bool fQuit; uint32_t i = 0; do { uTmpValue = puMem[i++]; fQuit = uTmpValue == uValueReg; } while (i < cLeftPage && !fQuit); /* Update the regs. */ RT_CONCAT(iemAImpl_cmp_u,OP_SIZE)((OP_TYPE *)&uValueReg, uTmpValue, &uEFlags); pCtx->ADDR_rCX = uCounterReg -= i; pCtx->ADDR_rDI = uAddrReg += i * cbIncr; pCtx->eflags.u = uEFlags; Assert(!!(uEFlags & X86_EFL_ZF) == fQuit); iemMemPageUnmap(pIemCpu, GCPhysMem, IEM_ACCESS_DATA_R, puMem, &PgLockMem); if (fQuit) break; /* If unaligned, we drop thru and do the page crossing access below. Otherwise, do the next page. */ if (!(uVirtAddr & (OP_SIZE - 1))) continue; if (uCounterReg == 0) break; cLeftPage = 0; } } /* * Fallback - slow processing till the end of the current page. * In the cross page boundrary case we will end up here with cLeftPage * as 0, we execute one loop then. */ do { OP_TYPE uTmpValue; rcStrict = RT_CONCAT(iemMemFetchDataU,OP_SIZE)(pIemCpu, &uTmpValue, X86_SREG_ES, uAddrReg); if (rcStrict != VINF_SUCCESS) return rcStrict; RT_CONCAT(iemAImpl_cmp_u,OP_SIZE)((OP_TYPE *)&uValueReg, uTmpValue, &uEFlags); pCtx->ADDR_rDI = uAddrReg += cbIncr; pCtx->ADDR_rCX = --uCounterReg; pCtx->eflags.u = uEFlags; cLeftPage--; } while ( (int32_t)cLeftPage > 0 && !(uEFlags & X86_EFL_ZF)); } while ( uCounterReg != 0 && !(uEFlags & X86_EFL_ZF)); /* * Done. */ iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } /** * Implements 'REP MOVS'. */ IEM_CIMPL_DEF_1(RT_CONCAT4(iemCImpl_rep_movs_op,OP_SIZE,_addr,ADDR_SIZE), uint8_t, iEffSeg) { PCPUMCTX pCtx = pIemCpu->CTX_SUFF(pCtx); /* * Setup. */ ADDR_TYPE uCounterReg = pCtx->ADDR_rCX; if (uCounterReg == 0) { iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } PCCPUMSELREGHID pSrcHid = iemSRegGetHid(pIemCpu, iEffSeg); uint64_t uSrcBase; VBOXSTRICTRC rcStrict = iemMemSegCheckReadAccessEx(pIemCpu, pSrcHid, iEffSeg, &uSrcBase); if (rcStrict != VINF_SUCCESS) return rcStrict; uint64_t uDstBase; rcStrict = iemMemSegCheckWriteAccessEx(pIemCpu, &pCtx->es, X86_SREG_ES, &uDstBase); if (rcStrict != VINF_SUCCESS) return rcStrict; int8_t const cbIncr = pCtx->eflags.Bits.u1DF ? -(OP_SIZE / 8) : (OP_SIZE / 8); ADDR_TYPE uSrcAddrReg = pCtx->ADDR_rSI; ADDR_TYPE uDstAddrReg = pCtx->ADDR_rDI; /* * Be careful with handle bypassing. */ if (pIemCpu->fBypassHandlers) { Log(("%s: declining because we're bypassing handlers\n", __FUNCTION__)); return VERR_IEM_ASPECT_NOT_IMPLEMENTED; } /* * If we're reading back what we write, we have to let the verfication code * to prevent a false positive. * Note! This doesn't take aliasing or wrapping into account - lazy bird. */ #ifdef IEM_VERIFICATION_MODE_FULL if ( IEM_VERIFICATION_ENABLED(pIemCpu) && (cbIncr > 0 ? uSrcAddrReg <= uDstAddrReg && uSrcAddrReg + cbIncr * uCounterReg > uDstAddrReg : uDstAddrReg <= uSrcAddrReg && uDstAddrReg + cbIncr * uCounterReg > uSrcAddrReg)) pIemCpu->fOverlappingMovs = true; #endif /* * The loop. */ do { /* * Do segmentation and virtual page stuff. */ ADDR2_TYPE uVirtSrcAddr = uSrcAddrReg + (ADDR2_TYPE)uSrcBase; ADDR2_TYPE uVirtDstAddr = uDstAddrReg + (ADDR2_TYPE)uDstBase; uint32_t cLeftSrcPage = (PAGE_SIZE - (uVirtSrcAddr & PAGE_OFFSET_MASK)) / (OP_SIZE / 8); if (cLeftSrcPage > uCounterReg) cLeftSrcPage = uCounterReg; uint32_t cLeftDstPage = (PAGE_SIZE - (uVirtDstAddr & PAGE_OFFSET_MASK)) / (OP_SIZE / 8); uint32_t cLeftPage = RT_MIN(cLeftSrcPage, cLeftDstPage); if ( cLeftPage > 0 /* can be null if unaligned, do one fallback round. */ && cbIncr > 0 /** @todo Implement reverse direction string ops. */ && ( IS_64_BIT_CODE(pIemCpu) || ( uSrcAddrReg < pSrcHid->u32Limit && uSrcAddrReg + (cLeftPage * (OP_SIZE / 8)) <= pSrcHid->u32Limit && uDstAddrReg < pCtx->es.u32Limit && uDstAddrReg + (cLeftPage * (OP_SIZE / 8)) <= pCtx->es.u32Limit) ) ) { RTGCPHYS GCPhysSrcMem; rcStrict = iemMemPageTranslateAndCheckAccess(pIemCpu, uVirtSrcAddr, IEM_ACCESS_DATA_R, &GCPhysSrcMem); if (rcStrict != VINF_SUCCESS) return rcStrict; RTGCPHYS GCPhysDstMem; rcStrict = iemMemPageTranslateAndCheckAccess(pIemCpu, uVirtDstAddr, IEM_ACCESS_DATA_W, &GCPhysDstMem); if (rcStrict != VINF_SUCCESS) return rcStrict; /* * If we can map the page without trouble, do a block processing * until the end of the current page. */ PGMPAGEMAPLOCK PgLockDstMem; OP_TYPE *puDstMem; rcStrict = iemMemPageMap(pIemCpu, GCPhysDstMem, IEM_ACCESS_DATA_W, (void **)&puDstMem, &PgLockDstMem); if (rcStrict == VINF_SUCCESS) { PGMPAGEMAPLOCK PgLockSrcMem; OP_TYPE const *puSrcMem; rcStrict = iemMemPageMap(pIemCpu, GCPhysSrcMem, IEM_ACCESS_DATA_R, (void **)&puSrcMem, &PgLockSrcMem); if (rcStrict == VINF_SUCCESS) { Assert( (GCPhysSrcMem >> PAGE_SHIFT) != (GCPhysDstMem >> PAGE_SHIFT) || ((uintptr_t)puSrcMem >> PAGE_SHIFT) == ((uintptr_t)puDstMem >> PAGE_SHIFT)); /* Perform the operation exactly (don't use memcpy to avoid having to consider how its implementation would affect any overlapping source and destination area). */ OP_TYPE const *puSrcCur = puSrcMem; OP_TYPE *puDstCur = puDstMem; uint32_t cTodo = cLeftPage; while (cTodo-- > 0) *puDstCur++ = *puSrcCur++; /* Update the registers. */ pCtx->ADDR_rSI = uSrcAddrReg += cLeftPage * cbIncr; pCtx->ADDR_rDI = uDstAddrReg += cLeftPage * cbIncr; pCtx->ADDR_rCX = uCounterReg -= cLeftPage; iemMemPageUnmap(pIemCpu, GCPhysSrcMem, IEM_ACCESS_DATA_R, puSrcMem, &PgLockSrcMem); iemMemPageUnmap(pIemCpu, GCPhysDstMem, IEM_ACCESS_DATA_W, puDstMem, &PgLockDstMem); continue; } iemMemPageUnmap(pIemCpu, GCPhysDstMem, IEM_ACCESS_DATA_W, puDstMem, &PgLockDstMem); } } /* * Fallback - slow processing till the end of the current page. * In the cross page boundrary case we will end up here with cLeftPage * as 0, we execute one loop then. */ do { OP_TYPE uValue; rcStrict = RT_CONCAT(iemMemFetchDataU,OP_SIZE)(pIemCpu, &uValue, iEffSeg, uSrcAddrReg); if (rcStrict != VINF_SUCCESS) return rcStrict; rcStrict = RT_CONCAT(iemMemStoreDataU,OP_SIZE)(pIemCpu, X86_SREG_ES, uDstAddrReg, uValue); if (rcStrict != VINF_SUCCESS) return rcStrict; pCtx->ADDR_rSI = uSrcAddrReg += cbIncr; pCtx->ADDR_rDI = uDstAddrReg += cbIncr; pCtx->ADDR_rCX = --uCounterReg; cLeftPage--; } while ((int32_t)cLeftPage > 0); } while (uCounterReg != 0); /* * Done. */ iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } /** * Implements 'REP STOS'. */ IEM_CIMPL_DEF_0(RT_CONCAT4(iemCImpl_stos_,OP_rAX,_m,ADDR_SIZE)) { PCPUMCTX pCtx = pIemCpu->CTX_SUFF(pCtx); /* * Setup. */ ADDR_TYPE uCounterReg = pCtx->ADDR_rCX; if (uCounterReg == 0) { iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } uint64_t uBaseAddr; VBOXSTRICTRC rcStrict = iemMemSegCheckWriteAccessEx(pIemCpu, &pCtx->es, X86_SREG_ES, &uBaseAddr); if (rcStrict != VINF_SUCCESS) return rcStrict; int8_t const cbIncr = pCtx->eflags.Bits.u1DF ? -(OP_SIZE / 8) : (OP_SIZE / 8); OP_TYPE const uValue = pCtx->OP_rAX; ADDR_TYPE uAddrReg = pCtx->ADDR_rDI; /* * Be careful with handle bypassing. */ /** @todo Permit doing a page if correctly aligned. */ if (pIemCpu->fBypassHandlers) { Log(("%s: declining because we're bypassing handlers\n", __FUNCTION__)); return VERR_IEM_ASPECT_NOT_IMPLEMENTED; } /* * The loop. */ do { /* * Do segmentation and virtual page stuff. */ ADDR2_TYPE uVirtAddr = uAddrReg + (ADDR2_TYPE)uBaseAddr; uint32_t cLeftPage = (PAGE_SIZE - (uVirtAddr & PAGE_OFFSET_MASK)) / (OP_SIZE / 8); if (cLeftPage > uCounterReg) cLeftPage = uCounterReg; if ( cLeftPage > 0 /* can be null if unaligned, do one fallback round. */ && cbIncr > 0 /** @todo Implement reverse direction string ops. */ && ( IS_64_BIT_CODE(pIemCpu) || ( uAddrReg < pCtx->es.u32Limit && uAddrReg + (cLeftPage * (OP_SIZE / 8)) <= pCtx->es.u32Limit) ) ) { RTGCPHYS GCPhysMem; rcStrict = iemMemPageTranslateAndCheckAccess(pIemCpu, uVirtAddr, IEM_ACCESS_DATA_W, &GCPhysMem); if (rcStrict != VINF_SUCCESS) return rcStrict; /* * If we can map the page without trouble, do a block processing * until the end of the current page. */ PGMPAGEMAPLOCK PgLockMem; OP_TYPE *puMem; rcStrict = iemMemPageMap(pIemCpu, GCPhysMem, IEM_ACCESS_DATA_W, (void **)&puMem, &PgLockMem); if (rcStrict == VINF_SUCCESS) { /* Update the regs first so we can loop on cLeftPage. */ pCtx->ADDR_rCX = uCounterReg -= cLeftPage; pCtx->ADDR_rDI = uAddrReg += cLeftPage * cbIncr; /* Do the memsetting. */ #if OP_SIZE == 8 memset(puMem, uValue, cLeftPage); /*#elif OP_SIZE == 32 ASMMemFill32(puMem, cLeftPage * (OP_SIZE / 8), uValue);*/ #else while (cLeftPage-- > 0) *puMem++ = uValue; #endif iemMemPageUnmap(pIemCpu, GCPhysMem, IEM_ACCESS_DATA_W, puMem, &PgLockMem); /* If unaligned, we drop thru and do the page crossing access below. Otherwise, do the next page. */ if (!(uVirtAddr & (OP_SIZE - 1))) continue; if (uCounterReg == 0) break; cLeftPage = 0; } } /* * Fallback - slow processing till the end of the current page. * In the cross page boundrary case we will end up here with cLeftPage * as 0, we execute one loop then. */ do { rcStrict = RT_CONCAT(iemMemStoreDataU,OP_SIZE)(pIemCpu, X86_SREG_ES, uAddrReg, uValue); if (rcStrict != VINF_SUCCESS) return rcStrict; pCtx->ADDR_rDI = uAddrReg += cbIncr; pCtx->ADDR_rCX = --uCounterReg; cLeftPage--; } while ((int32_t)cLeftPage > 0); } while (uCounterReg != 0); /* * Done. */ iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } /** * Implements 'REP LODS'. */ IEM_CIMPL_DEF_1(RT_CONCAT4(iemCImpl_lods_,OP_rAX,_m,ADDR_SIZE), int8_t, iEffSeg) { PCPUMCTX pCtx = pIemCpu->CTX_SUFF(pCtx); /* * Setup. */ ADDR_TYPE uCounterReg = pCtx->ADDR_rCX; if (uCounterReg == 0) { iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } PCCPUMSELREGHID pSrcHid = iemSRegGetHid(pIemCpu, iEffSeg); uint64_t uBaseAddr; VBOXSTRICTRC rcStrict = iemMemSegCheckReadAccessEx(pIemCpu, pSrcHid, iEffSeg, &uBaseAddr); if (rcStrict != VINF_SUCCESS) return rcStrict; int8_t const cbIncr = pCtx->eflags.Bits.u1DF ? -(OP_SIZE / 8) : (OP_SIZE / 8); ADDR_TYPE uAddrReg = pCtx->ADDR_rSI; /* * The loop. */ do { /* * Do segmentation and virtual page stuff. */ ADDR2_TYPE uVirtAddr = uAddrReg + (ADDR2_TYPE)uBaseAddr; uint32_t cLeftPage = (PAGE_SIZE - (uVirtAddr & PAGE_OFFSET_MASK)) / (OP_SIZE / 8); if (cLeftPage > uCounterReg) cLeftPage = uCounterReg; if ( cLeftPage > 0 /* can be null if unaligned, do one fallback round. */ && cbIncr > 0 /** @todo Implement reverse direction string ops. */ && ( IS_64_BIT_CODE(pIemCpu) || ( uAddrReg < pSrcHid->u32Limit && uAddrReg + (cLeftPage * (OP_SIZE / 8)) <= pSrcHid->u32Limit) ) ) { RTGCPHYS GCPhysMem; rcStrict = iemMemPageTranslateAndCheckAccess(pIemCpu, uVirtAddr, IEM_ACCESS_DATA_R, &GCPhysMem); if (rcStrict != VINF_SUCCESS) return rcStrict; /* * If we can map the page without trouble, we can get away with * just reading the last value on the page. */ PGMPAGEMAPLOCK PgLockMem; OP_TYPE const *puMem; rcStrict = iemMemPageMap(pIemCpu, GCPhysMem, IEM_ACCESS_DATA_R, (void **)&puMem, &PgLockMem); if (rcStrict == VINF_SUCCESS) { /* Only get the last byte, the rest doesn't matter in direct access mode. */ #if OP_SIZE == 32 pCtx->rax = puMem[cLeftPage - 1]; #else pCtx->OP_rAX = puMem[cLeftPage - 1]; #endif pCtx->ADDR_rCX = uCounterReg -= cLeftPage; pCtx->ADDR_rSI = uAddrReg += cLeftPage * cbIncr; iemMemPageUnmap(pIemCpu, GCPhysMem, IEM_ACCESS_DATA_R, puMem, &PgLockMem); /* If unaligned, we drop thru and do the page crossing access below. Otherwise, do the next page. */ if (!(uVirtAddr & (OP_SIZE - 1))) continue; if (uCounterReg == 0) break; cLeftPage = 0; } } /* * Fallback - slow processing till the end of the current page. * In the cross page boundrary case we will end up here with cLeftPage * as 0, we execute one loop then. */ do { OP_TYPE uTmpValue; rcStrict = RT_CONCAT(iemMemFetchDataU,OP_SIZE)(pIemCpu, &uTmpValue, iEffSeg, uAddrReg); if (rcStrict != VINF_SUCCESS) return rcStrict; #if OP_SIZE == 32 pCtx->rax = uTmpValue; #else pCtx->OP_rAX = uTmpValue; #endif pCtx->ADDR_rSI = uAddrReg += cbIncr; pCtx->ADDR_rCX = --uCounterReg; cLeftPage--; } while ((int32_t)cLeftPage > 0); if (rcStrict != VINF_SUCCESS) break; } while (uCounterReg != 0); /* * Done. */ iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } #if OP_SIZE != 64 /** * Implements 'INS' (no rep) */ IEM_CIMPL_DEF_1(RT_CONCAT4(iemCImpl_ins_op,OP_SIZE,_addr,ADDR_SIZE), bool, fIoChecked) { PVM pVM = IEMCPU_TO_VM(pIemCpu); PCPUMCTX pCtx = pIemCpu->CTX_SUFF(pCtx); VBOXSTRICTRC rcStrict; /* * Be careful with handle bypassing. */ if (pIemCpu->fBypassHandlers) { Log(("%s: declining because we're bypassing handlers\n", __FUNCTION__)); return VERR_IEM_ASPECT_NOT_IMPLEMENTED; } /* * ASSUMES the #GP for I/O permission is taken first, then any #GP for * segmentation and finally any #PF due to virtual address translation. * ASSUMES nothing is read from the I/O port before traps are taken. */ if (!fIoChecked) { rcStrict = iemHlpCheckPortIOPermission(pIemCpu, pCtx, pCtx->dx, OP_SIZE / 8); if (rcStrict != VINF_SUCCESS) return rcStrict; } OP_TYPE *puMem; rcStrict = iemMemMap(pIemCpu, (void **)&puMem, OP_SIZE / 8, X86_SREG_ES, pCtx->ADDR_rDI, IEM_ACCESS_DATA_W); if (rcStrict != VINF_SUCCESS) return rcStrict; uint32_t u32Value = 0; if (!IEM_VERIFICATION_ENABLED(pIemCpu)) rcStrict = IOMIOPortRead(pVM, IEMCPU_TO_VMCPU(pIemCpu), pCtx->dx, &u32Value, OP_SIZE / 8); else rcStrict = iemVerifyFakeIOPortRead(pIemCpu, pCtx->dx, &u32Value, OP_SIZE / 8); if (IOM_SUCCESS(rcStrict)) { VBOXSTRICTRC rcStrict2 = iemMemCommitAndUnmap(pIemCpu, puMem, IEM_ACCESS_DATA_W); if (RT_LIKELY(rcStrict2 == VINF_SUCCESS)) { if (!pCtx->eflags.Bits.u1DF) pCtx->ADDR_rDI += OP_SIZE / 8; else pCtx->ADDR_rDI -= OP_SIZE / 8; iemRegAddToRipAndClearRF(pIemCpu, cbInstr); } /* iemMemMap already check permissions, so this may only be real errors or access handlers medling. The access handler case is going to cause misbehavior if the instruction is re-interpreted or smth. So, we fail with an internal error here instead. */ else AssertLogRelMsgFailedReturn(("rcStrict2=%Rrc\n", VBOXSTRICTRC_VAL(rcStrict2)), VERR_IEM_IPE_1); } return rcStrict; } /** * Implements 'REP INS'. */ IEM_CIMPL_DEF_1(RT_CONCAT4(iemCImpl_rep_ins_op,OP_SIZE,_addr,ADDR_SIZE), bool, fIoChecked) { PVM pVM = IEMCPU_TO_VM(pIemCpu); PVMCPU pVCpu = IEMCPU_TO_VMCPU(pIemCpu); PCPUMCTX pCtx = pIemCpu->CTX_SUFF(pCtx); /* * Setup. */ uint16_t const u16Port = pCtx->dx; VBOXSTRICTRC rcStrict; if (!fIoChecked) { rcStrict = iemHlpCheckPortIOPermission(pIemCpu, pCtx, u16Port, OP_SIZE / 8); if (rcStrict != VINF_SUCCESS) return rcStrict; } ADDR_TYPE uCounterReg = pCtx->ADDR_rCX; if (uCounterReg == 0) { iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } uint64_t uBaseAddr; rcStrict = iemMemSegCheckWriteAccessEx(pIemCpu, &pCtx->es, X86_SREG_ES, &uBaseAddr); if (rcStrict != VINF_SUCCESS) return rcStrict; int8_t const cbIncr = pCtx->eflags.Bits.u1DF ? -(OP_SIZE / 8) : (OP_SIZE / 8); ADDR_TYPE uAddrReg = pCtx->ADDR_rDI; /* * Be careful with handle bypassing. */ if (pIemCpu->fBypassHandlers) { Log(("%s: declining because we're bypassing handlers\n", __FUNCTION__)); return VERR_IEM_ASPECT_NOT_IMPLEMENTED; } /* * The loop. */ do { /* * Do segmentation and virtual page stuff. */ ADDR2_TYPE uVirtAddr = uAddrReg + (ADDR2_TYPE)uBaseAddr; uint32_t cLeftPage = (PAGE_SIZE - (uVirtAddr & PAGE_OFFSET_MASK)) / (OP_SIZE / 8); if (cLeftPage > uCounterReg) cLeftPage = uCounterReg; if ( cLeftPage > 0 /* can be null if unaligned, do one fallback round. */ && cbIncr > 0 /** @todo Implement reverse direction string ops. */ && ( IS_64_BIT_CODE(pIemCpu) || ( uAddrReg < pCtx->es.u32Limit && uAddrReg + (cLeftPage * (OP_SIZE / 8)) <= pCtx->es.u32Limit) ) ) { RTGCPHYS GCPhysMem; rcStrict = iemMemPageTranslateAndCheckAccess(pIemCpu, uVirtAddr, IEM_ACCESS_DATA_W, &GCPhysMem); if (rcStrict != VINF_SUCCESS) return rcStrict; /* * If we can map the page without trouble, we would've liked to use * an string I/O method to do the work, but the current IOM * interface doesn't match our current approach. So, do a regular * loop instead. */ /** @todo Change the I/O manager interface to make use of * mapped buffers instead of leaving those bits to the * device implementation! */ PGMPAGEMAPLOCK PgLockMem; OP_TYPE *puMem; rcStrict = iemMemPageMap(pIemCpu, GCPhysMem, IEM_ACCESS_DATA_W, (void **)&puMem, &PgLockMem); if (rcStrict == VINF_SUCCESS) { uint32_t off = 0; while (off < cLeftPage) { uint32_t u32Value; if (!IEM_VERIFICATION_ENABLED(pIemCpu)) rcStrict = IOMIOPortRead(pVM, pVCpu, u16Port, &u32Value, OP_SIZE / 8); else rcStrict = iemVerifyFakeIOPortRead(pIemCpu, u16Port, &u32Value, OP_SIZE / 8); if (IOM_SUCCESS(rcStrict)) { puMem[off] = (OP_TYPE)u32Value; pCtx->ADDR_rDI = uAddrReg += cbIncr; pCtx->ADDR_rCX = --uCounterReg; } if (rcStrict != VINF_SUCCESS) { if (IOM_SUCCESS(rcStrict)) { rcStrict = iemSetPassUpStatus(pIemCpu, rcStrict); if (uCounterReg == 0) iemRegAddToRipAndClearRF(pIemCpu, cbInstr); } iemMemPageUnmap(pIemCpu, GCPhysMem, IEM_ACCESS_DATA_W, puMem, &PgLockMem); return rcStrict; } off++; } iemMemPageUnmap(pIemCpu, GCPhysMem, IEM_ACCESS_DATA_W, puMem, &PgLockMem); /* If unaligned, we drop thru and do the page crossing access below. Otherwise, do the next page. */ if (!(uVirtAddr & (OP_SIZE - 1))) continue; if (uCounterReg == 0) break; cLeftPage = 0; } } /* * Fallback - slow processing till the end of the current page. * In the cross page boundrary case we will end up here with cLeftPage * as 0, we execute one loop then. * * Note! We ASSUME the CPU will raise #PF or #GP before access the * I/O port, otherwise it wouldn't really be restartable. */ /** @todo investigate what the CPU actually does with \#PF/\#GP * during INS. */ do { OP_TYPE *puMem; rcStrict = iemMemMap(pIemCpu, (void **)&puMem, OP_SIZE / 8, X86_SREG_ES, uAddrReg, IEM_ACCESS_DATA_W); if (rcStrict != VINF_SUCCESS) return rcStrict; uint32_t u32Value = 0; if (!IEM_VERIFICATION_ENABLED(pIemCpu)) rcStrict = IOMIOPortRead(pVM, pVCpu, u16Port, &u32Value, OP_SIZE / 8); else rcStrict = iemVerifyFakeIOPortRead(pIemCpu, u16Port, &u32Value, OP_SIZE / 8); if (!IOM_SUCCESS(rcStrict)) return rcStrict; *puMem = (OP_TYPE)u32Value; VBOXSTRICTRC rcStrict2 = iemMemCommitAndUnmap(pIemCpu, puMem, IEM_ACCESS_DATA_W); AssertLogRelMsgReturn(rcStrict2 == VINF_SUCCESS, ("rcStrict2=%Rrc\n", VBOXSTRICTRC_VAL(rcStrict2)), VERR_IEM_IPE_1); /* See non-rep version. */ pCtx->ADDR_rDI = uAddrReg += cbIncr; pCtx->ADDR_rCX = --uCounterReg; cLeftPage--; if (rcStrict != VINF_SUCCESS) { if (uCounterReg == 0) iemRegAddToRipAndClearRF(pIemCpu, cbInstr); rcStrict = iemSetPassUpStatus(pIemCpu, rcStrict); return rcStrict; } } while ((int32_t)cLeftPage > 0); } while (uCounterReg != 0); /* * Done. */ iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } /** * Implements 'OUTS' (no rep) */ IEM_CIMPL_DEF_2(RT_CONCAT4(iemCImpl_outs_op,OP_SIZE,_addr,ADDR_SIZE), uint8_t, iEffSeg, bool, fIoChecked) { PVM pVM = IEMCPU_TO_VM(pIemCpu); PCPUMCTX pCtx = pIemCpu->CTX_SUFF(pCtx); VBOXSTRICTRC rcStrict; /* * ASSUMES the #GP for I/O permission is taken first, then any #GP for * segmentation and finally any #PF due to virtual address translation. * ASSUMES nothing is read from the I/O port before traps are taken. */ if (!fIoChecked) { rcStrict = iemHlpCheckPortIOPermission(pIemCpu, pCtx, pCtx->dx, OP_SIZE / 8); if (rcStrict != VINF_SUCCESS) return rcStrict; } OP_TYPE uValue; rcStrict = RT_CONCAT(iemMemFetchDataU,OP_SIZE)(pIemCpu, &uValue, iEffSeg, pCtx->ADDR_rSI); if (rcStrict == VINF_SUCCESS) { if (!IEM_VERIFICATION_ENABLED(pIemCpu)) rcStrict = IOMIOPortWrite(pVM, IEMCPU_TO_VMCPU(pIemCpu), pCtx->dx, uValue, OP_SIZE / 8); else rcStrict = iemVerifyFakeIOPortWrite(pIemCpu, pCtx->dx, uValue, OP_SIZE / 8); if (IOM_SUCCESS(rcStrict)) { if (!pCtx->eflags.Bits.u1DF) pCtx->ADDR_rSI += OP_SIZE / 8; else pCtx->ADDR_rSI -= OP_SIZE / 8; iemRegAddToRipAndClearRF(pIemCpu, cbInstr); if (rcStrict != VINF_SUCCESS) rcStrict = iemSetPassUpStatus(pIemCpu, rcStrict); } } return rcStrict; } /** * Implements 'REP OUTS'. */ IEM_CIMPL_DEF_2(RT_CONCAT4(iemCImpl_rep_outs_op,OP_SIZE,_addr,ADDR_SIZE), uint8_t, iEffSeg, bool, fIoChecked) { PVM pVM = IEMCPU_TO_VM(pIemCpu); PVMCPU pVCpu = IEMCPU_TO_VMCPU(pIemCpu); PCPUMCTX pCtx = pIemCpu->CTX_SUFF(pCtx); /* * Setup. */ uint16_t const u16Port = pCtx->dx; VBOXSTRICTRC rcStrict; if (!fIoChecked) { rcStrict = iemHlpCheckPortIOPermission(pIemCpu, pCtx, u16Port, OP_SIZE / 8); if (rcStrict != VINF_SUCCESS) return rcStrict; } ADDR_TYPE uCounterReg = pCtx->ADDR_rCX; if (uCounterReg == 0) { iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } PCCPUMSELREGHID pHid = iemSRegGetHid(pIemCpu, iEffSeg); uint64_t uBaseAddr; rcStrict = iemMemSegCheckReadAccessEx(pIemCpu, pHid, iEffSeg, &uBaseAddr); if (rcStrict != VINF_SUCCESS) return rcStrict; int8_t const cbIncr = pCtx->eflags.Bits.u1DF ? -(OP_SIZE / 8) : (OP_SIZE / 8); ADDR_TYPE uAddrReg = pCtx->ADDR_rSI; /* * The loop. */ do { /* * Do segmentation and virtual page stuff. */ ADDR2_TYPE uVirtAddr = uAddrReg + (ADDR2_TYPE)uBaseAddr; uint32_t cLeftPage = (PAGE_SIZE - (uVirtAddr & PAGE_OFFSET_MASK)) / (OP_SIZE / 8); if (cLeftPage > uCounterReg) cLeftPage = uCounterReg; if ( cLeftPage > 0 /* can be null if unaligned, do one fallback round. */ && cbIncr > 0 /** @todo Implement reverse direction string ops. */ && ( IS_64_BIT_CODE(pIemCpu) || ( uAddrReg < pHid->u32Limit && uAddrReg + (cLeftPage * (OP_SIZE / 8)) <= pHid->u32Limit) ) ) { RTGCPHYS GCPhysMem; rcStrict = iemMemPageTranslateAndCheckAccess(pIemCpu, uVirtAddr, IEM_ACCESS_DATA_R, &GCPhysMem); if (rcStrict != VINF_SUCCESS) return rcStrict; /* * If we can map the page without trouble, we would've liked to use * an string I/O method to do the work, but the current IOM * interface doesn't match our current approach. So, do a regular * loop instead. */ /** @todo Change the I/O manager interface to make use of * mapped buffers instead of leaving those bits to the * device implementation? */ PGMPAGEMAPLOCK PgLockMem; OP_TYPE const *puMem; rcStrict = iemMemPageMap(pIemCpu, GCPhysMem, IEM_ACCESS_DATA_R, (void **)&puMem, &PgLockMem); if (rcStrict == VINF_SUCCESS) { uint32_t off = 0; while (off < cLeftPage) { uint32_t u32Value = *puMem++; if (!IEM_VERIFICATION_ENABLED(pIemCpu)) rcStrict = IOMIOPortWrite(pVM, pVCpu, u16Port, u32Value, OP_SIZE / 8); else rcStrict = iemVerifyFakeIOPortWrite(pIemCpu, u16Port, u32Value, OP_SIZE / 8); if (IOM_SUCCESS(rcStrict)) { pCtx->ADDR_rSI = uAddrReg += cbIncr; pCtx->ADDR_rCX = --uCounterReg; } if (rcStrict != VINF_SUCCESS) { if (IOM_SUCCESS(rcStrict)) { rcStrict = iemSetPassUpStatus(pIemCpu, rcStrict); if (uCounterReg == 0) iemRegAddToRipAndClearRF(pIemCpu, cbInstr); } iemMemPageUnmap(pIemCpu, GCPhysMem, IEM_ACCESS_DATA_R, puMem, &PgLockMem); return rcStrict; } off++; } iemMemPageUnmap(pIemCpu, GCPhysMem, IEM_ACCESS_DATA_R, puMem, &PgLockMem); /* If unaligned, we drop thru and do the page crossing access below. Otherwise, do the next page. */ if (!(uVirtAddr & (OP_SIZE - 1))) continue; if (uCounterReg == 0) break; cLeftPage = 0; } } /* * Fallback - slow processing till the end of the current page. * In the cross page boundrary case we will end up here with cLeftPage * as 0, we execute one loop then. * * Note! We ASSUME the CPU will raise #PF or #GP before access the * I/O port, otherwise it wouldn't really be restartable. */ /** @todo investigate what the CPU actually does with \#PF/\#GP * during INS. */ do { OP_TYPE uValue; rcStrict = RT_CONCAT(iemMemFetchDataU,OP_SIZE)(pIemCpu, &uValue, iEffSeg, uAddrReg); if (rcStrict != VINF_SUCCESS) return rcStrict; if (!IEM_VERIFICATION_ENABLED(pIemCpu)) rcStrict = IOMIOPortWrite(pVM, pVCpu, u16Port, uValue, OP_SIZE / 8); else rcStrict = iemVerifyFakeIOPortWrite(pIemCpu, u16Port, uValue, OP_SIZE / 8); if (IOM_SUCCESS(rcStrict)) { pCtx->ADDR_rSI = uAddrReg += cbIncr; pCtx->ADDR_rCX = --uCounterReg; cLeftPage--; } if (rcStrict != VINF_SUCCESS) { if (IOM_SUCCESS(rcStrict)) { if (uCounterReg == 0) iemRegAddToRipAndClearRF(pIemCpu, cbInstr); rcStrict = iemSetPassUpStatus(pIemCpu, rcStrict); } return rcStrict; } } while ((int32_t)cLeftPage > 0); } while (uCounterReg != 0); /* * Done. */ iemRegAddToRipAndClearRF(pIemCpu, cbInstr); return VINF_SUCCESS; } #endif /* OP_SIZE != 64-bit */ #undef OP_rAX #undef OP_SIZE #undef ADDR_SIZE #undef ADDR_rDI #undef ADDR_rSI #undef ADDR_rCX #undef ADDR_rIP #undef ADDR2_TYPE #undef ADDR_TYPE #undef ADDR2_TYPE #undef IS_64_BIT_CODE