From: Lai Jiangshan laijs@linux.alibaba.com
stable inclusion from stable-v5.10.84 commit 2015ffa3a4c2b058627558ef39b3cce6fe422bf4 bugzilla: 186030 https://gitee.com/openeuler/kernel/issues/I4QV2F
Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=...
--------------------------------
[ Upstream commit 5c8f6a2e316efebb3ba93d8c1af258155dcf5632 ]
In the native case, PER_CPU_VAR(cpu_tss_rw + TSS_sp0) is the trampoline stack. But XEN pv doesn't use trampoline stack, so PER_CPU_VAR(cpu_tss_rw + TSS_sp0) is also the kernel stack.
In that case, source and destination stacks are identical, which means that reusing swapgs_restore_regs_and_return_to_usermode() in XEN pv would cause %rsp to move up to the top of the kernel stack and leave the IRET frame below %rsp.
This is dangerous as it can be corrupted if #NMI / #MC hit as either of these events occurring in the middle of the stack pushing would clobber data on the (original) stack.
And, with XEN pv, swapgs_restore_regs_and_return_to_usermode() pushing the IRET frame on to the original address is useless and error-prone when there is any future attempt to modify the code.
[ bp: Massage commit message. ]
Fixes: 7f2590a110b8 ("x86/entry/64: Use a per-CPU trampoline stack for IDT entries") Signed-off-by: Lai Jiangshan laijs@linux.alibaba.com Signed-off-by: Borislav Petkov bp@suse.de Reviewed-by: Boris Ostrovsky boris.ostrovsky@oracle.com Link: https://lkml.kernel.org/r/20211126101209.8613-4-jiangshanlai@gmail.com Signed-off-by: Sasha Levin sashal@kernel.org Signed-off-by: Chen Jun chenjun102@huawei.com Signed-off-by: Zheng Zengkai zhengzengkai@huawei.com --- arch/x86/entry/entry_64.S | 4 ++++ arch/x86/xen/xen-asm.S | 20 ++++++++++++++++++++ 2 files changed, 24 insertions(+)
diff --git a/arch/x86/entry/entry_64.S b/arch/x86/entry/entry_64.S index a806d68b9699..de541ea2788e 100644 --- a/arch/x86/entry/entry_64.S +++ b/arch/x86/entry/entry_64.S @@ -575,6 +575,10 @@ SYM_INNER_LABEL(swapgs_restore_regs_and_return_to_usermode, SYM_L_GLOBAL) ud2 1: #endif +#ifdef CONFIG_XEN_PV + ALTERNATIVE "", "jmp xenpv_restore_regs_and_return_to_usermode", X86_FEATURE_XENPV +#endif + POP_REGS pop_rdi=0
/* diff --git a/arch/x86/xen/xen-asm.S b/arch/x86/xen/xen-asm.S index 53cf8aa35032..011ec649f388 100644 --- a/arch/x86/xen/xen-asm.S +++ b/arch/x86/xen/xen-asm.S @@ -19,6 +19,7 @@
#include <linux/init.h> #include <linux/linkage.h> +#include <../entry/calling.h>
/* * Enable events. This clears the event mask and tests the pending @@ -235,6 +236,25 @@ SYM_CODE_START(xen_sysret64) jmp hypercall_iret SYM_CODE_END(xen_sysret64)
+/* + * XEN pv doesn't use trampoline stack, PER_CPU_VAR(cpu_tss_rw + TSS_sp0) is + * also the kernel stack. Reusing swapgs_restore_regs_and_return_to_usermode() + * in XEN pv would cause %rsp to move up to the top of the kernel stack and + * leave the IRET frame below %rsp, which is dangerous to be corrupted if #NMI + * interrupts. And swapgs_restore_regs_and_return_to_usermode() pushing the IRET + * frame at the same address is useless. + */ +SYM_CODE_START(xenpv_restore_regs_and_return_to_usermode) + UNWIND_HINT_REGS + POP_REGS + + /* stackleak_erase() can work safely on the kernel stack. */ + STACKLEAK_ERASE_NOCLOBBER + + addq $8, %rsp /* skip regs->orig_ax */ + jmp xen_iret +SYM_CODE_END(xenpv_restore_regs_and_return_to_usermode) + /* * Xen handles syscall callbacks much like ordinary exceptions, which * means we have: