All of lore.kernel.org
 help / color / mirror / Atom feed
From: Andrew Cooper <andrew.cooper3@citrix.com>
To: Xen-devel <xen-devel@lists.xenproject.org>
Cc: "Andrew Cooper" <andrew.cooper3@citrix.com>,
	"Jan Beulich" <JBeulich@suse.com>,
	"Roger Pau Monné" <roger.pau@citrix.com>, "Wei Liu" <wl@xen.org>
Subject: [PATCH 64/65] x86/efi: Disable CET-IBT around Runtime Services calls
Date: Fri, 26 Nov 2021 12:34:45 +0000	[thread overview]
Message-ID: <20211126123446.32324-65-andrew.cooper3@citrix.com> (raw)
In-Reply-To: <20211126123446.32324-1-andrew.cooper3@citrix.com>

At least one TigerLake NUC has UEFI firmware which isn't CET-IBT compatible.
Read under a function pointer to see whether an endbr64 instruction is
present, and use this as a heuristic.

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
---
CC: Jan Beulich <JBeulich@suse.com>
CC: Roger Pau Monné <roger.pau@citrix.com>
CC: Wei Liu <wl@xen.org>

This was disappointing to discover.  I've pestered some folk and maybe
something will improve in due course, but it remains an open question how best
to discover that Runtime Services are CET-IBT compatible.
---
 xen/arch/x86/efi/stub.c  |  2 ++
 xen/common/efi/boot.c    |  6 ++++++
 xen/common/efi/runtime.c | 17 +++++++++++++++++
 xen/include/xen/efi.h    |  1 +
 4 files changed, 26 insertions(+)

diff --git a/xen/arch/x86/efi/stub.c b/xen/arch/x86/efi/stub.c
index 998493262641..5e44913e52db 100644
--- a/xen/arch/x86/efi/stub.c
+++ b/xen/arch/x86/efi/stub.c
@@ -11,6 +11,8 @@
 #include <efi/efidevp.h>
 #include <efi/efiapi.h>
 
+bool __initdata efi_no_cet_ibt;
+
 /*
  * Here we are in EFI stub. EFI calls are not supported due to lack
  * of relevant functionality in compiler and/or linker.
diff --git a/xen/common/efi/boot.c b/xen/common/efi/boot.c
index f5af71837d5a..2c7f86f4f534 100644
--- a/xen/common/efi/boot.c
+++ b/xen/common/efi/boot.c
@@ -735,6 +735,12 @@ static void __init efi_init(EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE *SystemTabl
 
     StdOut = SystemTable->ConOut;
     StdErr = SystemTable->StdErr ?: StdOut;
+
+    /*
+     * Heuristic.  Look under an arbitrary function pointer to see if UEFI was
+     * compiled with CET-IBT support.  Experimentally some are not.
+     */
+    efi_no_cet_ibt = memcmp(efi_rs->GetTime, "\xf3\x0f\x1e\xfa", 4) != 0;
 }
 
 static void __init efi_console_set_mode(void)
diff --git a/xen/common/efi/runtime.c b/xen/common/efi/runtime.c
index d2fdc28df3e0..ef54863542db 100644
--- a/xen/common/efi/runtime.c
+++ b/xen/common/efi/runtime.c
@@ -21,6 +21,7 @@ struct efi_rs_state {
   * don't strictly need that.
   */
  unsigned long __aligned(32) cr3;
+    unsigned long msr_s_cet;
 #endif
 };
 
@@ -61,6 +62,7 @@ UINTN __read_mostly efi_apple_properties_len;
 
 /* Bit field representing available EFI features/properties. */
 unsigned int efi_flags;
+bool __read_mostly efi_no_cet_ibt;
 
 struct efi __read_mostly efi = {
 	.acpi   = EFI_INVALID_TABLE_ADDR,
@@ -113,6 +115,17 @@ struct efi_rs_state efi_rs_enter(void)
 
     switch_cr3_cr4(mfn_to_maddr(efi_l4_mfn), read_cr4());
 
+    /*
+     * If UEFI doesn't appear to be CET-IBT compatible, stash and clobber
+     * ENDBR_EN.  Always read the current CET setting, because CET-SS isn't
+     * configured until very late on the BSP.
+     */
+    if ( cpu_has_xen_ibt && efi_no_cet_ibt )
+    {
+        rdmsrl(MSR_S_CET, state.msr_s_cet);
+        wrmsrl(MSR_S_CET, state.msr_s_cet & ~CET_ENDBR_EN);
+    }
+
     return state;
 }
 
@@ -122,6 +135,10 @@ void efi_rs_leave(struct efi_rs_state *state)
 
     if ( !state->cr3 )
         return;
+
+    if ( state->msr_s_cet )
+        wrmsrl(MSR_S_CET, state->msr_s_cet);
+
     switch_cr3_cr4(state->cr3, read_cr4());
     if ( is_pv_vcpu(curr) && !is_idle_vcpu(curr) )
     {
diff --git a/xen/include/xen/efi.h b/xen/include/xen/efi.h
index 94a7e547f97b..8c14f7f18718 100644
--- a/xen/include/xen/efi.h
+++ b/xen/include/xen/efi.h
@@ -30,6 +30,7 @@ union compat_pf_efi_info;
 
 struct xenpf_efi_runtime_call;
 struct compat_pf_efi_runtime_call;
+extern bool efi_no_cet_ibt;
 
 bool efi_enabled(unsigned int feature);
 void efi_init_memory(void);
-- 
2.11.0



  parent reply	other threads:[~2021-11-26 13:08 UTC|newest]

Thread overview: 129+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-11-26 12:33 [PATCH 00/65] x86: Support for CET Indirect Branch Tracking Andrew Cooper
2021-11-26 12:33 ` [PATCH 01/65] x86: Introduce support for CET-IBT Andrew Cooper
2021-11-26 14:10   ` Jan Beulich
2021-11-26 15:21     ` Andrew Cooper
2021-11-29  9:21       ` Jan Beulich
2021-12-10 14:20         ` Andrew Cooper
2021-12-13  7:32           ` Jan Beulich
2021-11-29  9:27   ` Jan Beulich
2021-11-29 10:27     ` Andrew Cooper
2021-11-26 12:33 ` [PATCH 02/65] x86/pv-shim: don't modify hypercall table Andrew Cooper
2021-11-26 12:33 ` [PATCH 03/65] xen/x86: don't use hypercall table for calling compat hypercalls Andrew Cooper
2021-11-26 12:33 ` [PATCH 04/65] x86/hypercall: Annotate fnptr targets Andrew Cooper
2021-11-26 14:21   ` Jan Beulich
2021-11-26 14:28     ` Andrew Cooper
2021-11-29  9:38       ` Jan Beulich
2021-11-29 10:00         ` Jan Beulich
2021-11-26 12:33 ` [PATCH 05/65] xen: Annotate fnptr targets from custom_param() Andrew Cooper
2021-11-26 12:33 ` [PATCH 06/65] xen: Annotate fnptr targets from __initcall() Andrew Cooper
2021-11-26 12:33 ` [PATCH 07/65] xen: Annotate fnptr targets from notifier callbacks Andrew Cooper
2021-11-26 12:33 ` [PATCH 08/65] xen: Annotate fnptr targets from acpi_table_parse() Andrew Cooper
2021-12-06  8:36   ` Jan Beulich
2021-12-10 14:44     ` Andrew Cooper
2021-12-13  7:46       ` Jan Beulich
2021-11-26 12:33 ` [PATCH 09/65] xen: Annotate fnptr targets from continue_hypercall_on_cpu() Andrew Cooper
2021-11-26 12:33 ` [PATCH 10/65] xen: Annotate fnptr targets from init_timer() Andrew Cooper
2021-11-26 12:33 ` [PATCH 11/65] xen: Annotate fnptr targets from call_rcu() Andrew Cooper
2021-11-26 12:33 ` [PATCH 12/65] xen: Annotate fnptr targets from IPIs Andrew Cooper
2021-11-26 12:33 ` [PATCH 13/65] xen: Annotate fnptr targets from open_softirq() Andrew Cooper
2021-11-26 12:33 ` [PATCH 14/65] xsm/flask: Annotate fnptr targets in the security server Andrew Cooper
2021-12-03 10:44   ` Daniel P. Smith
2021-11-26 12:33 ` [PATCH 15/65] xsm: Annotate fnptr targets Andrew Cooper
2021-12-03 10:46   ` Daniel P. Smith
2021-11-26 12:33 ` [PATCH 16/65] xen/sched: " Andrew Cooper
2021-11-26 12:33 ` [PATCH 17/65] xen/evtchn: " Andrew Cooper
2021-11-29 13:51   ` Jan Beulich
2021-11-26 12:33 ` [PATCH 18/65] xen/hypfs: " Andrew Cooper
2021-12-02 11:35   ` Juergen Gross
2021-11-26 12:34 ` [PATCH 19/65] xen/tasklet: " Andrew Cooper
2021-11-29 13:56   ` Jan Beulich
2021-11-26 12:34 ` [PATCH 20/65] xen/keyhandler: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 21/65] xen/vpci: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 22/65] xen/decompress: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 23/65] xen/iommu: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 24/65] xen/video: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 25/65] xen/console: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 26/65] xen/misc: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 27/65] x86: Annotate fnptr targets from request_irq() Andrew Cooper
2021-11-26 12:34 ` [PATCH 28/65] x86/hvm: Annotate fnptr targets from hvm_funcs Andrew Cooper
2021-11-29 14:14   ` Jan Beulich
2021-11-26 12:34 ` [PATCH 29/65] x86/hvm: Annotate fnptr targets from device emulation Andrew Cooper
2021-11-26 12:34 ` [PATCH 30/65] x86/emul: Annotate fnptr targets Andrew Cooper
2021-11-29 14:22   ` Jan Beulich
2021-11-26 12:34 ` [PATCH 31/65] x86/ucode: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 32/65] x86/power: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 33/65] x86/apic: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 34/65] x86/nmi: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 35/65] x86/mtrr: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 36/65] x86/idle: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 37/65] x86/quirks: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 38/65] x86/hvmsave: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 39/65] x86/mce: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 40/65] x86/pmu: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 41/65] x86/cpu: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 42/65] x86/guest: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 43/65] x86/logdirty: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 44/65] x86/shadow: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 45/65] x86/hap: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 46/65] x86/p2m: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 47/65] x86/irq: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 48/65] x86/aepi: " Andrew Cooper
2021-11-29 14:33   ` Jan Beulich
2021-11-26 12:34 ` [PATCH 49/65] x86/psr: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 50/65] x86/dpci: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 51/65] x86/pt: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 52/65] x86/time: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 53/65] x86/misc: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 54/65] x86/stack: " Andrew Cooper
2021-11-29 14:41   ` Jan Beulich
2021-11-26 12:34 ` [PATCH 55/65] x86/bugframe: " Andrew Cooper
2021-11-26 12:34 ` [PATCH 56/65] x86: Use control flow typechecking where possible Andrew Cooper
2021-11-26 12:34 ` [PATCH 57/65] x86/setup: Read CR4 earlier in __start_xen() Andrew Cooper
2021-12-03 12:58   ` Jan Beulich
2021-11-26 12:34 ` [PATCH 58/65] x86/alternatives: Clear CR4.CET when clearing CR0.WP Andrew Cooper
2021-12-03 12:59   ` Jan Beulich
2021-11-26 12:34 ` [PATCH 59/65] x86/traps: Rework write_stub_trampoline() to not hardcode the jmp Andrew Cooper
2021-12-03 13:17   ` Jan Beulich
2021-12-03 13:59     ` Andrew Cooper
2021-12-03 14:03       ` Jan Beulich
2021-11-26 12:34 ` [PATCH 60/65] x86/emul: Update emulation stubs to be CET-IBT compatible Andrew Cooper
2021-11-26 12:34 ` [PATCH 61/65] x86/entry: Make syscall/sysenter entrypoints " Andrew Cooper
2021-11-26 12:34 ` [PATCH 62/65] x86/entry: Make IDT " Andrew Cooper
2021-12-03 13:23   ` Jan Beulich
2021-12-03 14:12     ` Andrew Cooper
2021-12-03 13:32   ` Jan Beulich
2021-12-03 15:30     ` Andrew Cooper
2021-12-06  9:42       ` Jan Beulich
2021-12-06 11:38         ` Andrew Cooper
2021-12-06 12:27           ` Jan Beulich
2021-11-26 12:34 ` [PATCH 63/65] x86/setup: Rework MSR_S_CET handling for CET-IBT Andrew Cooper
2021-12-06 10:49   ` Jan Beulich
2021-12-10 16:19     ` Andrew Cooper
2021-12-10 16:47       ` Jan Beulich
2021-11-26 12:34 ` Andrew Cooper [this message]
2021-11-26 12:34 ` [PATCH 65/65] x86: Enable CET Indirect Branch Tracking Andrew Cooper
2021-12-06 11:24   ` Jan Beulich
2021-11-26 12:48 ` [PATCH 00/65] x86: Support for " Jan Beulich
2021-11-26 13:13   ` Andrew Cooper
2021-11-26 13:21     ` Andrew Cooper
2021-11-26 13:22     ` Jan Beulich
2021-11-26 23:49       ` Andrew Cooper
2021-11-29  9:18         ` Jan Beulich
2021-11-26 16:33 ` [PATCH 59.5/65] x86: Introduce helpers/checks for endbr64 instructions Andrew Cooper
2021-11-26 18:26   ` Marek Marczykowski-Górecki
2021-11-26 18:29     ` Andrew Cooper
2021-12-03 13:59   ` Jan Beulich
2021-12-03 14:10     ` Andrew Cooper
2021-12-06  8:04       ` Jan Beulich
2021-11-26 16:37 ` [PATCH v1.1 60/65] x86/emul: Update emulation stubs to be CET-IBT compatible Andrew Cooper
2021-12-03 13:39   ` Jan Beulich
2021-11-26 16:37 ` [PATCH v1.1 61/65] x86/entry: Make syscall/sysenter entrypoints " Andrew Cooper
2021-12-03 13:48   ` Jan Beulich
2021-11-26 16:38 ` [PATCH v1.1 64/65] x86/efi: Disable CET-IBT around Runtime Services calls Andrew Cooper
2021-12-06 11:06   ` Jan Beulich
2021-12-10 17:16     ` Andrew Cooper
2021-12-13  7:52       ` Jan Beulich
2021-11-29 14:44 ` [PATCH 00/65] x86: Support for CET Indirect Branch Tracking Jan Beulich
2021-11-29 15:09   ` Andrew Cooper
2021-11-29 16:17     ` Jan Beulich
2021-12-02 10:31   ` Dario Faggioli

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20211126123446.32324-65-andrew.cooper3@citrix.com \
    --to=andrew.cooper3@citrix.com \
    --cc=JBeulich@suse.com \
    --cc=roger.pau@citrix.com \
    --cc=wl@xen.org \
    --cc=xen-devel@lists.xenproject.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.