Currently attempt to do MMIO from userspace in TDX guest leads to
warning about unexpect #VE and SIGSEGV being delivered to the process.
Enlightened userspace might choose to deal with MMIO on their own if
kernel doesn't emulate it.
Handle EPT_VIOLATION exit reason for userspace and deliver SIGBUS
instead of SIGSEV. SIGBUS is more appropriate for MMIO situation.
Signed-off-by: Kirill A. Shutemov <[email protected]>
---
arch/x86/coco/tdx/tdx.c | 19 ++++++++++++++-----
1 file changed, 14 insertions(+), 5 deletions(-)
diff --git a/arch/x86/coco/tdx/tdx.c b/arch/x86/coco/tdx/tdx.c
index c1cb90369915..d2aa93cebf5a 100644
--- a/arch/x86/coco/tdx/tdx.c
+++ b/arch/x86/coco/tdx/tdx.c
@@ -7,6 +7,7 @@
#include <linux/cpufeature.h>
#include <linux/export.h>
#include <linux/io.h>
+#include <linux/sched/signal.h>
#include <asm/coco.h>
#include <asm/tdx.h>
#include <asm/vmx.h>
@@ -630,6 +631,11 @@ void tdx_get_ve_info(struct ve_info *ve)
ve->instr_info = upper_32_bits(args.r10);
}
+static inline bool is_private_gpa(u64 gpa)
+{
+ return gpa == cc_mkenc(gpa);
+}
+
/*
* Handle the user initiated #VE.
*
@@ -641,17 +647,20 @@ static int virt_exception_user(struct pt_regs *regs, struct ve_info *ve)
switch (ve->exit_reason) {
case EXIT_REASON_CPUID:
return handle_cpuid(regs, ve);
+ case EXIT_REASON_EPT_VIOLATION:
+ if (is_private_gpa(ve->gpa))
+ panic("Unexpected EPT-violation on private memory.");
+
+ force_sig_fault(SIGBUS, BUS_ADRERR, (void __user *)ve->gla);
+
+ /* Return 0 to avoid incrementing RIP */
+ return 0;
default:
pr_warn("Unexpected #VE: %lld\n", ve->exit_reason);
return -EIO;
}
}
-static inline bool is_private_gpa(u64 gpa)
-{
- return gpa == cc_mkenc(gpa);
-}
-
/*
* Handle the kernel #VE.
*
--
2.43.0
On 5/21/24 12:35 AM, Kirill A. Shutemov wrote:
> Currently attempt to do MMIO from userspace in TDX guest leads to
> warning about unexpect #VE and SIGSEGV being delivered to the process.
>
> Enlightened userspace might choose to deal with MMIO on their own if
> kernel doesn't emulate it.
Any specific use cases ? Like who is using it?
> Handle EPT_VIOLATION exit reason for userspace and deliver SIGBUS
> instead of SIGSEV. SIGBUS is more appropriate for MMIO situation.
>
> Signed-off-by: Kirill A. Shutemov <[email protected]>
> ---
Code looks good to me.
Reviewed-by: Kuppuswamy Sathyanarayanan <[email protected]>
> arch/x86/coco/tdx/tdx.c | 19 ++++++++++++++-----
> 1 file changed, 14 insertions(+), 5 deletions(-)
>
> diff --git a/arch/x86/coco/tdx/tdx.c b/arch/x86/coco/tdx/tdx.c
> index c1cb90369915..d2aa93cebf5a 100644
> --- a/arch/x86/coco/tdx/tdx.c
> +++ b/arch/x86/coco/tdx/tdx.c
> @@ -7,6 +7,7 @@
> #include <linux/cpufeature.h>
> #include <linux/export.h>
> #include <linux/io.h>
> +#include <linux/sched/signal.h>
> #include <asm/coco.h>
> #include <asm/tdx.h>
> #include <asm/vmx.h>
> @@ -630,6 +631,11 @@ void tdx_get_ve_info(struct ve_info *ve)
> ve->instr_info = upper_32_bits(args.r10);
> }
>
> +static inline bool is_private_gpa(u64 gpa)
> +{
> + return gpa == cc_mkenc(gpa);
> +}
> +
> /*
> * Handle the user initiated #VE.
> *
> @@ -641,17 +647,20 @@ static int virt_exception_user(struct pt_regs *regs, struct ve_info *ve)
> switch (ve->exit_reason) {
> case EXIT_REASON_CPUID:
> return handle_cpuid(regs, ve);
> + case EXIT_REASON_EPT_VIOLATION:
> + if (is_private_gpa(ve->gpa))
> + panic("Unexpected EPT-violation on private memory.");
> +
> + force_sig_fault(SIGBUS, BUS_ADRERR, (void __user *)ve->gla);
> +
> + /* Return 0 to avoid incrementing RIP */
> + return 0;
> default:
> pr_warn("Unexpected #VE: %lld\n", ve->exit_reason);
> return -EIO;
> }
> }
>
> -static inline bool is_private_gpa(u64 gpa)
> -{
> - return gpa == cc_mkenc(gpa);
> -}
> -
> /*
> * Handle the kernel #VE.
> *
--
Sathyanarayanan Kuppuswamy
Linux Kernel Developer
On Tue, May 21, 2024 at 06:35:49AM -0700, Kuppuswamy Sathyanarayanan wrote:
>
> On 5/21/24 12:35 AM, Kirill A. Shutemov wrote:
> > Currently attempt to do MMIO from userspace in TDX guest leads to
> > warning about unexpect #VE and SIGSEGV being delivered to the process.
> >
> > Enlightened userspace might choose to deal with MMIO on their own if
> > kernel doesn't emulate it.
>
> Any specific use cases ? Like who is using it?
Microsoft folks wanted it. Chris, Dexuan, John, any comments?
But it is generally right thing to do. SIGBUS is right signal to deliver.
--
Kiryl Shutsemau / Kirill A. Shutemov
We use this to handle MMIO issued by userspace that the kernel does not handle in a #VE, for devices assigned to a TDX VM.
Chris
-----Original Message-----
From: Kirill A. Shutemov <[email protected]>
Sent: Thursday, May 23, 2024 3:15 AM
To: Kuppuswamy Sathyanarayanan <[email protected]>
Cc: Dave Hansen <[email protected]>; Thomas Gleixner <[email protected]>; Ingo Molnar <[email protected]>; Borislav Petkov <[email protected]>; [email protected]; H. Peter Anvin <[email protected]>; [email protected]; [email protected]; Chris Oo <[email protected]>; Dexuan Cui <[email protected]>; John Starks <[email protected]>
Subject: [EXTERNAL] Re: [PATCH] x86/tdx: Generate SIGBUS on userspace MMIO
On Tue, May 21, 2024 at 06:35:49AM -0700, Kuppuswamy Sathyanarayanan wrote:
>
> On 5/21/24 12:35 AM, Kirill A. Shutemov wrote:
> > Currently attempt to do MMIO from userspace in TDX guest leads to
> > warning about unexpect #VE and SIGSEGV being delivered to the process.
> >
> > Enlightened userspace might choose to deal with MMIO on their own if
> > kernel doesn't emulate it.
>
> Any specific use cases ? Like who is using it?
Microsoft folks wanted it. Chris, Dexuan, John, any comments?
But it is generally right thing to do. SIGBUS is right signal to deliver.
--
Kiryl Shutsemau / Kirill A. Shutemov