[PATCH] x86/sev: Fix kernel crash due to late update to read-only ghcb_version

[Ingo Molnar]

* Ashwin Dayanand Kamat <kashwindayan@xxxxxxxxxx> wrote:

> From: Ashwin Dayanand Kamat <ashwin.kamat@xxxxxxxxxxxx>
> 
> kernel crash was observed because of page fault, while running
> cpuhotplug ltp testcases on SEV-ES enabled systems. The crash was
> observed during hotplug after the CPU was offlined and the process
> was migrated to different cpu. setup_ghcb() is called again which
> tries to update ghcb_version in sev_es_negotiate_protocol(). Ideally this
> is a read_only variable which is initialised during booting.
> This results in pagefault.

Applied to tip:x86/urgent, thanks.

Tom: I've added your Suggested-by and Acked-by, which appeared to be the 
case given the v1 discussion, let me know if that's not accurate.

I've also tidied up the changelog - final version attached below.

Thanks,

	Ingo

============>
From: Ashwin Dayanand Kamat <ashwin.kamat@xxxxxxxxxxxx>
Date: Wed, 29 Nov 2023 16:10:29 +0530
Subject: [PATCH] x86/sev: Fix kernel crash due to late update to read-only ghcb_version

A write-access violation page fault kernel crash was observed while running
cpuhotplug LTP testcases on SEV-ES enabled systems. The crash was
observed during hotplug, after the CPU was offlined and the process
was migrated to different CPU. setup_ghcb() is called again which
tries to update ghcb_version in sev_es_negotiate_protocol(). Ideally this
is a read_only variable which is initialised during booting.

Trying to write it results in a pagefault:

  BUG: unable to handle page fault for address: ffffffffba556e70
  #PF: supervisor write access in kernel mode
  #PF: error_code(0x0003) - permissions violation
  [ ...]
  Call Trace:
   <TASK>
   ? __die_body.cold+0x1a/0x1f
   ? __die+0x2a/0x35
   ? page_fault_oops+0x10c/0x270
   ? setup_ghcb+0x71/0x100
   ? __x86_return_thunk+0x5/0x6
   ? search_exception_tables+0x60/0x70
   ? __x86_return_thunk+0x5/0x6
   ? fixup_exception+0x27/0x320
   ? kernelmode_fixup_or_oops+0xa2/0x120
   ? __bad_area_nosemaphore+0x16a/0x1b0
   ? kernel_exc_vmm_communication+0x60/0xb0
   ? bad_area_nosemaphore+0x16/0x20
   ? do_kern_addr_fault+0x7a/0x90
   ? exc_page_fault+0xbd/0x160
   ? asm_exc_page_fault+0x27/0x30
   ? setup_ghcb+0x71/0x100
   ? setup_ghcb+0xe/0x100
   cpu_init_exception_handling+0x1b9/0x1f0

The fix is to call sev_es_negotiate_protocol() only in the BSP boot phase,
and it only needs to be done once in any case.

[ mingo: Refined the changelog. ]

Fixes: 95d33bfaa3e1 ("x86/sev: Register GHCB memory when SEV-SNP is active")
Suggested-by: Tom Lendacky <thomas.lendacky@xxxxxxx>
Co-developed-by: Bo Gan <bo.gan@xxxxxxxxxxxx>
Signed-off-by: Bo Gan <bo.gan@xxxxxxxxxxxx>
Signed-off-by: Ashwin Dayanand Kamat <ashwin.kamat@xxxxxxxxxxxx>
Signed-off-by: Ingo Molnar <mingo@xxxxxxxxxx>
Acked-by: Tom Lendacky <thomas.lendacky@xxxxxxx>
Link: https://lore.kernel.org/r/1701254429-18250-1-git-send-email-kashwindayan@xxxxxxxxxx
---
 arch/x86/kernel/sev.c | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/arch/x86/kernel/sev.c b/arch/x86/kernel/sev.c
index 70472eebe719..c67285824e82 100644
--- a/arch/x86/kernel/sev.c
+++ b/arch/x86/kernel/sev.c
@@ -1234,10 +1234,6 @@ void setup_ghcb(void)
 	if (!cc_platform_has(CC_ATTR_GUEST_STATE_ENCRYPT))
 		return;
 
-	/* First make sure the hypervisor talks a supported protocol. */
-	if (!sev_es_negotiate_protocol())
-		sev_es_terminate(SEV_TERM_SET_GEN, GHCB_SEV_ES_GEN_REQ);
-
 	/*
 	 * Check whether the runtime #VC exception handler is active. It uses
 	 * the per-CPU GHCB page which is set up by sev_es_init_vc_handling().
@@ -1254,6 +1250,13 @@ void setup_ghcb(void)
 		return;
 	}
 
+	/*
+	 * Make sure the hypervisor talks a supported protocol.
+	 * This gets called only in the BSP boot phase.
+	 */
+	if (!sev_es_negotiate_protocol())
+		sev_es_terminate(SEV_TERM_SET_GEN, GHCB_SEV_ES_GEN_REQ);
+
 	/*
 	 * Clear the boot_ghcb. The first exception comes in before the bss
 	 * section is cleared.