提交 318f5a2a 编写于 作者: A Andy Lutomirski 提交者: H. Peter Anvin

x86-64: Add user_64bit_mode paravirt op

Three places in the kernel assume that the only long mode CPL 3
selector is __USER_CS.  This is not true on Xen -- Xen's sysretq
changes cs to the magic value 0xe033.

Two of the places are corner cases, but as of "x86-64: Improve
vsyscall emulation CS and RIP handling"
(c9712944), vsyscalls will segfault
if called with Xen's extra CS selector.  This causes a panic when
older init builds die.

It seems impossible to make Xen use __USER_CS reliably without
taking a performance hit on every system call, so this fixes the
tests instead with a new paravirt op.  It's a little ugly because
ptrace.h can't include paravirt.h.
Signed-off-by: NAndy Lutomirski <luto@mit.edu>
Link: http://lkml.kernel.org/r/f4fcb3947340d9e96ce1054a432f183f9da9db83.1312378163.git.luto@mit.eduReported-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: NH. Peter Anvin <hpa@linux.intel.com>
上级 5d5791af
...@@ -27,8 +27,8 @@ static inline void fill_ldt(struct desc_struct *desc, const struct user_desc *in ...@@ -27,8 +27,8 @@ static inline void fill_ldt(struct desc_struct *desc, const struct user_desc *in
desc->base2 = (info->base_addr & 0xff000000) >> 24; desc->base2 = (info->base_addr & 0xff000000) >> 24;
/* /*
* Don't allow setting of the lm bit. It is useless anyway * Don't allow setting of the lm bit. It would confuse
* because 64bit system calls require __USER_CS: * user_64bit_mode and would get overridden by sysret anyway.
*/ */
desc->l = 0; desc->l = 0;
} }
......
...@@ -41,6 +41,7 @@ ...@@ -41,6 +41,7 @@
#include <asm/desc_defs.h> #include <asm/desc_defs.h>
#include <asm/kmap_types.h> #include <asm/kmap_types.h>
#include <asm/pgtable_types.h>
struct page; struct page;
struct thread_struct; struct thread_struct;
...@@ -63,6 +64,11 @@ struct paravirt_callee_save { ...@@ -63,6 +64,11 @@ struct paravirt_callee_save {
struct pv_info { struct pv_info {
unsigned int kernel_rpl; unsigned int kernel_rpl;
int shared_kernel_pmd; int shared_kernel_pmd;
#ifdef CONFIG_X86_64
u16 extra_user_64bit_cs; /* __USER_CS if none */
#endif
int paravirt_enabled; int paravirt_enabled;
const char *name; const char *name;
}; };
......
...@@ -131,6 +131,9 @@ struct pt_regs { ...@@ -131,6 +131,9 @@ struct pt_regs {
#ifdef __KERNEL__ #ifdef __KERNEL__
#include <linux/init.h> #include <linux/init.h>
#ifdef CONFIG_PARAVIRT
#include <asm/paravirt_types.h>
#endif
struct cpuinfo_x86; struct cpuinfo_x86;
struct task_struct; struct task_struct;
...@@ -187,6 +190,22 @@ static inline int v8086_mode(struct pt_regs *regs) ...@@ -187,6 +190,22 @@ static inline int v8086_mode(struct pt_regs *regs)
#endif #endif
} }
#ifdef CONFIG_X86_64
static inline bool user_64bit_mode(struct pt_regs *regs)
{
#ifndef CONFIG_PARAVIRT
/*
* On non-paravirt systems, this is the only long mode CPL 3
* selector. We do not allow long mode selectors in the LDT.
*/
return regs->cs == __USER_CS;
#else
/* Headers are too twisted for this to go in paravirt.h. */
return regs->cs == __USER_CS || regs->cs == pv_info.extra_user_64bit_cs;
#endif
}
#endif
/* /*
* X86_32 CPUs don't save ss and esp if the CPU is already in kernel mode * X86_32 CPUs don't save ss and esp if the CPU is already in kernel mode
* when it traps. The previous stack will be directly underneath the saved * when it traps. The previous stack will be directly underneath the saved
......
...@@ -299,6 +299,10 @@ struct pv_info pv_info = { ...@@ -299,6 +299,10 @@ struct pv_info pv_info = {
.paravirt_enabled = 0, .paravirt_enabled = 0,
.kernel_rpl = 0, .kernel_rpl = 0,
.shared_kernel_pmd = 1, /* Only used when CONFIG_X86_PAE is set */ .shared_kernel_pmd = 1, /* Only used when CONFIG_X86_PAE is set */
#ifdef CONFIG_X86_64
.extra_user_64bit_cs = __USER_CS,
#endif
}; };
struct pv_init_ops pv_init_ops = { struct pv_init_ops pv_init_ops = {
......
...@@ -74,7 +74,7 @@ static int is_setting_trap_flag(struct task_struct *child, struct pt_regs *regs) ...@@ -74,7 +74,7 @@ static int is_setting_trap_flag(struct task_struct *child, struct pt_regs *regs)
#ifdef CONFIG_X86_64 #ifdef CONFIG_X86_64
case 0x40 ... 0x4f: case 0x40 ... 0x4f:
if (regs->cs != __USER_CS) if (!user_64bit_mode(regs))
/* 32-bit mode: register increment */ /* 32-bit mode: register increment */
return 0; return 0;
/* 64-bit mode: REX prefix */ /* 64-bit mode: REX prefix */
......
...@@ -127,11 +127,7 @@ void dotraplinkage do_emulate_vsyscall(struct pt_regs *regs, long error_code) ...@@ -127,11 +127,7 @@ void dotraplinkage do_emulate_vsyscall(struct pt_regs *regs, long error_code)
local_irq_enable(); local_irq_enable();
/* if (!user_64bit_mode(regs)) {
* Real 64-bit user mode code has cs == __USER_CS. Anything else
* is bogus.
*/
if (regs->cs != __USER_CS) {
/* /*
* If we trapped from kernel mode, we might as well OOPS now * If we trapped from kernel mode, we might as well OOPS now
* instead of returning to some random address and OOPSing * instead of returning to some random address and OOPSing
......
...@@ -105,7 +105,7 @@ check_prefetch_opcode(struct pt_regs *regs, unsigned char *instr, ...@@ -105,7 +105,7 @@ check_prefetch_opcode(struct pt_regs *regs, unsigned char *instr,
* but for now it's good enough to assume that long * but for now it's good enough to assume that long
* mode only uses well known segments or kernel. * mode only uses well known segments or kernel.
*/ */
return (!user_mode(regs)) || (regs->cs == __USER_CS); return (!user_mode(regs) || user_64bit_mode(regs));
#endif #endif
case 0x60: case 0x60:
/* 0x64 thru 0x67 are valid prefixes in all modes. */ /* 0x64 thru 0x67 are valid prefixes in all modes. */
......
...@@ -937,6 +937,10 @@ static const struct pv_info xen_info __initconst = { ...@@ -937,6 +937,10 @@ static const struct pv_info xen_info __initconst = {
.paravirt_enabled = 1, .paravirt_enabled = 1,
.shared_kernel_pmd = 0, .shared_kernel_pmd = 0,
#ifdef CONFIG_X86_64
.extra_user_64bit_cs = FLAT_USER_CS64,
#endif
.name = "Xen", .name = "Xen",
}; };
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册