From 3606189fa3da6afcad0cbbc9b91e94f1f158da5a Mon Sep 17 00:00:00 2001 From: Paolo Bonzini Date: Mon, 27 Oct 2014 14:40:49 +0100 Subject: KVM: emulator: fix error code for __linearize The error code for #GP and #SS is zero when the segment is used to access an operand or an instruction. It is only non-zero when a segment register is being loaded; for limit checks this means cases such as: * for #GP, when RIP is beyond the limit on a far call (before the first instruction is executed). We do not implement this check, but it would be in em_jmp_far/em_call_far. * for #SS, if the new stack overflows during an inter-privilege-level call to a non-conforming code segment. We do not implement stack switching at all. So use an error code of zero. Reviewed-by: Nadav Amit Signed-off-by: Paolo Bonzini --- arch/x86/kvm/emulate.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86/kvm/emulate.c') diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c index 749f9fa38254..8aa66068712f 100644 --- a/arch/x86/kvm/emulate.c +++ b/arch/x86/kvm/emulate.c @@ -711,9 +711,9 @@ static int __linearize(struct x86_emulate_ctxt *ctxt, return X86EMUL_CONTINUE; bad: if (addr.seg == VCPU_SREG_SS) - return emulate_ss(ctxt, sel); + return emulate_ss(ctxt, 0); else - return emulate_gp(ctxt, sel); + return emulate_gp(ctxt, 0); } static int linearize(struct x86_emulate_ctxt *ctxt, -- cgit v1.2.2 From fd56e1546a5f734290cbedd2b81c518850736511 Mon Sep 17 00:00:00 2001 From: Paolo Bonzini Date: Mon, 27 Oct 2014 14:40:39 +0100 Subject: KVM: emulator: fix execution close to the segment limit Emulation of code that is 14 bytes to the segment limit or closer (e.g. RIP = 0xFFFFFFF2 after reset) is broken because we try to read as many as 15 bytes from the beginning of the instruction, and __linearize fails when the passed (address, size) pair reaches out of the segment. To fix this, let __linearize return the maximum accessible size (clamped to 2^32-1) for usage in __do_insn_fetch_bytes, and avoid the limit check by passing zero for the desired size. For expand-down segments, __linearize is performing a redundant check. (u32)(addr.ea + size - 1) <= lim can only happen if addr.ea is close to 4GB; in this case, addr.ea + size - 1 will also fail the check against the upper bound of the segment (which is provided by the D/B bit). After eliminating the redundant check, it is simple to compute the *max_size for expand-down segments too. Now that the limit check is done in __do_insn_fetch_bytes, we want to inject a general protection fault there if size < op_size (like __linearize would have done), instead of just aborting. This fixes booting Tiano Core from emulated flash with EPT disabled. Cc: stable@vger.kernel.org Fixes: 719d5a9b2487e0562f178f61e323c3dc18a8b200 Reported-by: Borislav Petkov Tested-by: Borislav Petkov Signed-off-by: Paolo Bonzini --- arch/x86/kvm/emulate.c | 43 +++++++++++++++++++++++++++++++++---------- 1 file changed, 33 insertions(+), 10 deletions(-) (limited to 'arch/x86/kvm/emulate.c') diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c index 8aa66068712f..52a96270b560 100644 --- a/arch/x86/kvm/emulate.c +++ b/arch/x86/kvm/emulate.c @@ -641,7 +641,8 @@ static bool insn_aligned(struct x86_emulate_ctxt *ctxt, unsigned size) static int __linearize(struct x86_emulate_ctxt *ctxt, struct segmented_address addr, - unsigned size, bool write, bool fetch, + unsigned *max_size, unsigned size, + bool write, bool fetch, ulong *linear) { struct desc_struct desc; @@ -652,10 +653,15 @@ static int __linearize(struct x86_emulate_ctxt *ctxt, unsigned cpl; la = seg_base(ctxt, addr.seg) + addr.ea; + *max_size = 0; switch (ctxt->mode) { case X86EMUL_MODE_PROT64: if (((signed long)la << 16) >> 16 != la) return emulate_gp(ctxt, 0); + + *max_size = min_t(u64, ~0u, (1ull << 48) - la); + if (size > *max_size) + goto bad; break; default: usable = ctxt->ops->get_segment(ctxt, &sel, &desc, NULL, @@ -673,20 +679,25 @@ static int __linearize(struct x86_emulate_ctxt *ctxt, if ((ctxt->mode == X86EMUL_MODE_REAL) && !fetch && (ctxt->d & NoBigReal)) { /* la is between zero and 0xffff */ - if (la > 0xffff || (u32)(la + size - 1) > 0xffff) + if (la > 0xffff) goto bad; + *max_size = 0x10000 - la; } else if ((desc.type & 8) || !(desc.type & 4)) { /* expand-up segment */ - if (addr.ea > lim || (u32)(addr.ea + size - 1) > lim) + if (addr.ea > lim) goto bad; + *max_size = min_t(u64, ~0u, (u64)lim + 1 - addr.ea); } else { /* expand-down segment */ - if (addr.ea <= lim || (u32)(addr.ea + size - 1) <= lim) + if (addr.ea <= lim) goto bad; lim = desc.d ? 0xffffffff : 0xffff; - if (addr.ea > lim || (u32)(addr.ea + size - 1) > lim) + if (addr.ea > lim) goto bad; + *max_size = min_t(u64, ~0u, (u64)lim + 1 - addr.ea); } + if (size > *max_size) + goto bad; cpl = ctxt->ops->cpl(ctxt); if (!(desc.type & 8)) { /* data segment */ @@ -721,7 +732,8 @@ static int linearize(struct x86_emulate_ctxt *ctxt, unsigned size, bool write, ulong *linear) { - return __linearize(ctxt, addr, size, write, false, linear); + unsigned max_size; + return __linearize(ctxt, addr, &max_size, size, write, false, linear); } @@ -746,17 +758,27 @@ static int segmented_read_std(struct x86_emulate_ctxt *ctxt, static int __do_insn_fetch_bytes(struct x86_emulate_ctxt *ctxt, int op_size) { int rc; - unsigned size; + unsigned size, max_size; unsigned long linear; int cur_size = ctxt->fetch.end - ctxt->fetch.data; struct segmented_address addr = { .seg = VCPU_SREG_CS, .ea = ctxt->eip + cur_size }; - size = 15UL ^ cur_size; - rc = __linearize(ctxt, addr, size, false, true, &linear); + /* + * We do not know exactly how many bytes will be needed, and + * __linearize is expensive, so fetch as much as possible. We + * just have to avoid going beyond the 15 byte limit, the end + * of the segment, or the end of the page. + * + * __linearize is called with size 0 so that it does not do any + * boundary check itself. Instead, we use max_size to check + * against op_size. + */ + rc = __linearize(ctxt, addr, &max_size, 0, false, true, &linear); if (unlikely(rc != X86EMUL_CONTINUE)) return rc; + size = min_t(unsigned, 15UL ^ cur_size, max_size); size = min_t(unsigned, size, PAGE_SIZE - offset_in_page(linear)); /* @@ -766,7 +788,8 @@ static int __do_insn_fetch_bytes(struct x86_emulate_ctxt *ctxt, int op_size) * still, we must have hit the 15-byte boundary. */ if (unlikely(size < op_size)) - return X86EMUL_UNHANDLEABLE; + return emulate_gp(ctxt, 0); + rc = ctxt->ops->fetch(ctxt, linear, ctxt->fetch.end, size, &ctxt->exception); if (unlikely(rc != X86EMUL_CONTINUE)) -- cgit v1.2.2 From 7e46dddd6f6cd5dbf3c7bd04a7e75d19475ac9f2 Mon Sep 17 00:00:00 2001 From: Nadav Amit Date: Tue, 28 Oct 2014 00:03:43 +0200 Subject: KVM: x86: Fix far-jump to non-canonical check Commit d1442d85cc30 ("KVM: x86: Handle errors when RIP is set during far jumps") introduced a bug that caused the fix to be incomplete. Due to incorrect evaluation, far jump to segment with L bit cleared (i.e., 32-bit segment) and RIP with any of the high bits set (i.e, RIP[63:32] != 0) set may not trigger #GP. As we know, this imposes a security problem. In addition, the condition for two warnings was incorrect. Fixes: d1442d85cc30ea75f7d399474ca738e0bc96f715 Reported-by: Dan Carpenter Signed-off-by: Nadav Amit [Add #ifdef CONFIG_X86_64 to avoid complaints of undefined behavior. - Paolo] Signed-off-by: Paolo Bonzini --- arch/x86/kvm/emulate.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'arch/x86/kvm/emulate.c') diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c index 52a96270b560..5edf088ca51e 100644 --- a/arch/x86/kvm/emulate.c +++ b/arch/x86/kvm/emulate.c @@ -574,12 +574,14 @@ static inline int assign_eip_far(struct x86_emulate_ctxt *ctxt, ulong dst, case 4: ctxt->_eip = (u32)dst; break; +#ifdef CONFIG_X86_64 case 8: if ((cs_l && is_noncanonical_address(dst)) || - (!cs_l && (dst & ~(u32)-1))) + (!cs_l && (dst >> 32) != 0)) return emulate_gp(ctxt, 0); ctxt->_eip = dst; break; +#endif default: WARN(1, "unsupported eip assignment size\n"); } @@ -2035,7 +2037,7 @@ static int em_jmp_far(struct x86_emulate_ctxt *ctxt) rc = assign_eip_far(ctxt, ctxt->src.val, new_desc.l); if (rc != X86EMUL_CONTINUE) { - WARN_ON(!ctxt->mode != X86EMUL_MODE_PROT64); + WARN_ON(ctxt->mode != X86EMUL_MODE_PROT64); /* assigning eip failed; restore the old cs */ ops->set_segment(ctxt, old_sel, &old_desc, 0, VCPU_SREG_CS); return rc; @@ -2132,7 +2134,7 @@ static int em_ret_far(struct x86_emulate_ctxt *ctxt) return rc; rc = assign_eip_far(ctxt, eip, new_desc.l); if (rc != X86EMUL_CONTINUE) { - WARN_ON(!ctxt->mode != X86EMUL_MODE_PROT64); + WARN_ON(ctxt->mode != X86EMUL_MODE_PROT64); ops->set_segment(ctxt, old_cs, &old_desc, 0, VCPU_SREG_CS); } return rc; -- cgit v1.2.2