[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <4483ef98b020b9d4cdafef95ee84b18eaccea61a.1430159990.git.jslaby@suse.cz>
Date: Mon, 27 Apr 2015 23:32:50 +0200
From: Jiri Slaby <jslaby@...e.cz>
To: stable@...r.kernel.org
Cc: linux-kernel@...r.kernel.org, Nadav Amit <namit@...technion.ac.il>,
Paolo Bonzini <pbonzini@...hat.com>,
Zhiqiang Zhang <zhangzhiqiang.zhang@...wei.com>,
Jiri Slaby <jslaby@...e.cz>
Subject: [PATCH 3.12 79/83] KVM: x86: SYSENTER emulation is broken
From: Nadav Amit <namit@...technion.ac.il>
3.12-stable review patch. If anyone has any objections, please let me know.
===============
commit f3747379accba8e95d70cec0eae0582c8c182050 upstream.
SYSENTER emulation is broken in several ways:
1. It misses the case of 16-bit code segments completely (CVE-2015-0239).
2. MSR_IA32_SYSENTER_CS is checked in 64-bit mode incorrectly (bits 0 and 1 can
still be set without causing #GP).
3. MSR_IA32_SYSENTER_EIP and MSR_IA32_SYSENTER_ESP are not masked in
legacy-mode.
4. There is some unneeded code.
Fix it.
Signed-off-by: Nadav Amit <namit@...technion.ac.il>
Signed-off-by: Paolo Bonzini <pbonzini@...hat.com>
[zhangzhiqiang: backport to 3.10:
- adjust context
- in 3.10 context "ctxt->eflags &= ~(EFLG_VM | EFLG_IF | EFLG_RF)" is replaced by
"ctxt->eflags &= ~(EFLG_VM | EFLG_IF)" in upstream, which was changed by another commit.
- After the above adjustments, becomes same to the original patch:
https://github.com/torvalds/linux/commit/f3747379accba8e95d70cec0eae0582c8c182050
]
Signed-off-by: Zhiqiang Zhang <zhangzhiqiang.zhang@...wei.com>
Signed-off-by: Jiri Slaby <jslaby@...e.cz>
---
arch/x86/kvm/emulate.c | 27 ++++++++-------------------
1 file changed, 8 insertions(+), 19 deletions(-)
diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c
index c412bab82d1f..8216f484398f 100644
--- a/arch/x86/kvm/emulate.c
+++ b/arch/x86/kvm/emulate.c
@@ -2292,7 +2292,7 @@ static int em_sysenter(struct x86_emulate_ctxt *ctxt)
* Not recognized on AMD in compat mode (but is recognized in legacy
* mode).
*/
- if ((ctxt->mode == X86EMUL_MODE_PROT32) && (efer & EFER_LMA)
+ if ((ctxt->mode != X86EMUL_MODE_PROT64) && (efer & EFER_LMA)
&& !vendor_intel(ctxt))
return emulate_ud(ctxt);
@@ -2305,25 +2305,13 @@ static int em_sysenter(struct x86_emulate_ctxt *ctxt)
setup_syscalls_segments(ctxt, &cs, &ss);
ops->get_msr(ctxt, MSR_IA32_SYSENTER_CS, &msr_data);
- switch (ctxt->mode) {
- case X86EMUL_MODE_PROT32:
- if ((msr_data & 0xfffc) == 0x0)
- return emulate_gp(ctxt, 0);
- break;
- case X86EMUL_MODE_PROT64:
- if (msr_data == 0x0)
- return emulate_gp(ctxt, 0);
- break;
- default:
- break;
- }
+ if ((msr_data & 0xfffc) == 0x0)
+ return emulate_gp(ctxt, 0);
ctxt->eflags &= ~(EFLG_VM | EFLG_IF | EFLG_RF);
- cs_sel = (u16)msr_data;
- cs_sel &= ~SELECTOR_RPL_MASK;
+ cs_sel = (u16)msr_data & ~SELECTOR_RPL_MASK;
ss_sel = cs_sel + 8;
- ss_sel &= ~SELECTOR_RPL_MASK;
- if (ctxt->mode == X86EMUL_MODE_PROT64 || (efer & EFER_LMA)) {
+ if (efer & EFER_LMA) {
cs.d = 0;
cs.l = 1;
}
@@ -2332,10 +2320,11 @@ static int em_sysenter(struct x86_emulate_ctxt *ctxt)
ops->set_segment(ctxt, ss_sel, &ss, 0, VCPU_SREG_SS);
ops->get_msr(ctxt, MSR_IA32_SYSENTER_EIP, &msr_data);
- ctxt->_eip = msr_data;
+ ctxt->_eip = (efer & EFER_LMA) ? msr_data : (u32)msr_data;
ops->get_msr(ctxt, MSR_IA32_SYSENTER_ESP, &msr_data);
- *reg_write(ctxt, VCPU_REGS_RSP) = msr_data;
+ *reg_write(ctxt, VCPU_REGS_RSP) = (efer & EFER_LMA) ? msr_data :
+ (u32)msr_data;
return X86EMUL_CONTINUE;
}
--
2.3.5
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists