Touching privately mapped GPA that is not properly converted to private
with MapGPA and accepted leads to an unrecoverable exit to VMM.
load_unaligned_zeropad() can touch memory that is not owned by the
caller, but just happened to next after the owned memory.
This load_unaligned_zeropad() behaviour makes it important when kernel
asks VMM to convert a GPA from shared to private or back. Kernel must
never have a page mapped into direct mapping (and aliases) as private
when the GPA is already converted to shared or when GPA is not yet
converted to private.
load_unaligned_zeropad() can touch memory that is not owned by the
caller, but just happens to be next after the owned memory. This
load_unaligned_zeropad() behavior makes it important when the kernel
asks VMM to convert a GPA from shared to private or back. The kernel
must never have a page mapped into direct mapping (and aliases) as
private when the GPA is already converted to shared or when the GPA is
not yet converted to private.
guest.enc_status_change_prepare() is called before adjusting direct
mapping and therefore is responsible for converting the memory to
private.
guest.enc_status_change_finish() is called after adjusting direct
mapping and it converts the memory to shared.
It is okay to have a shared mapping of memory that is not properly
converted. handle_mmio() knows how to deal with load_unaligned_zeropad()
stepping on it.
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Fixes: 7dbde7631629 ("x86/mm/cpa: Add support for TDX shared memory")
Reviewed-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
---
arch/x86/coco/tdx/tdx.c | 64 +++++++++++++++++++++++++++++++++++++++--
1 file changed, 61 insertions(+), 3 deletions(-)
diff --git a/arch/x86/coco/tdx/tdx.c b/arch/x86/coco/tdx/tdx.c
index e146b599260f..f6213a10de3a 100644
--- a/arch/x86/coco/tdx/tdx.c
+++ b/arch/x86/coco/tdx/tdx.c
@@ -840,6 +840,30 @@ static bool tdx_enc_status_changed(unsigned long vaddr, int numpages, bool enc)
return true;
}
+static bool tdx_enc_status_change_prepare(unsigned long vaddr, int numpages,
+ bool enc)
+{
+ /*
+ * Only handle shared->private conversion here.
+ * See the comment in tdx_early_init().
+ */
+ if (enc)
+ return tdx_enc_status_changed(vaddr, numpages, enc);
+ return true;
+}
+
+static bool tdx_enc_status_change_finish(unsigned long vaddr, int numpages,
+ bool enc)
+{
+ /*
+ * Only handle private->shared conversion here.
+ * See the comment in tdx_early_init().
+ */
+ if (!enc)
+ return tdx_enc_status_changed(vaddr, numpages, enc);
+ return true;
+}
+
void __init tdx_early_init(void)
{
u64 cc_mask;
@@ -867,9 +891,43 @@ void __init tdx_early_init(void)
*/
physical_mask &= cc_mask - 1;
- x86_platform.guest.enc_cache_flush_required = tdx_cache_flush_required;
- x86_platform.guest.enc_tlb_flush_required = tdx_tlb_flush_required;
- x86_platform.guest.enc_status_change_finish = tdx_enc_status_changed;
+ /*
+ * Touching privately mapped GPA that is not properly converted to
+ * private with MapGPA and accepted leads to an unrecoverable exit
+ * to VMM.
+ *
+ * load_unaligned_zeropad() can touch memory that is not owned by the
+ * caller, but just happened to next after the owned memory.
+ * This load_unaligned_zeropad() behaviour makes it important when
+ * kernel asks VMM to convert a GPA from shared to private or back.
+ * Kernel must never have a page mapped into direct mapping (and
+ * aliases) as private when the GPA is already converted to shared or
+ * when GPA is not yet converted to private.
+ *
+ * load_unaligned_zeropad() can touch memory that is not owned by the
+ * caller, but just happens to be next after the owned memory. This
+ * load_unaligned_zeropad() behavior makes it important when the kernel
+ * asks VMM to convert a GPA from shared to private or back. The kernel
+ * must never have a page mapped into direct mapping (and aliases) as
+ * private when the GPA is already converted to shared or when the GPA
+ * is not yet converted to private.
+ *
+ * guest.enc_status_change_prepare() is called before adjusting direct
+ * mapping and therefore is responsible for converting the memory to
+ * private.
+ *
+ * guest.enc_status_change_finish() is called after adjusting direct
+ * mapping and it converts the memory to shared.
+ *
+ * It is okay to have a shared mapping of memory that is not properly
+ * converted. handle_mmio() knows how to deal with
+ * load_unaligned_zeropad() stepping on it.
+ */
+ x86_platform.guest.enc_status_change_prepare = tdx_enc_status_change_prepare;
+ x86_platform.guest.enc_status_change_finish = tdx_enc_status_change_finish;
+
+ x86_platform.guest.enc_cache_flush_required = tdx_cache_flush_required;
+ x86_platform.guest.enc_tlb_flush_required = tdx_tlb_flush_required;
pr_info("Guest detected\n");
}
--
2.39.3
On 6/6/23 02:56, Kirill A. Shutemov wrote: > load_unaligned_zeropad() can touch memory that is not owned by the > caller, but just happened to next after the owned memory. > This load_unaligned_zeropad() behaviour makes it important when kernel > asks VMM to convert a GPA from shared to private or back. Kernel must > never have a page mapped into direct mapping (and aliases) as private > when the GPA is already converted to shared or when GPA is not yet > converted to private. > > load_unaligned_zeropad() can touch memory that is not owned by the > caller, but just happens to be next after the owned memory. This > load_unaligned_zeropad() behavior makes it important when the kernel > asks VMM to convert a GPA from shared to private or back. The kernel > must never have a page mapped into direct mapping (and aliases) as > private when the GPA is already converted to shared or when the GPA is > not yet converted to private. Heh, that must be really important info to have it in the changelog twice! I'll fix it up when I apply it.
On Tue, Jun 06, 2023 at 11:14:29AM -0700, Dave Hansen wrote: > On 6/6/23 02:56, Kirill A. Shutemov wrote: > > load_unaligned_zeropad() can touch memory that is not owned by the > > caller, but just happened to next after the owned memory. > > This load_unaligned_zeropad() behaviour makes it important when kernel > > asks VMM to convert a GPA from shared to private or back. Kernel must > > never have a page mapped into direct mapping (and aliases) as private > > when the GPA is already converted to shared or when GPA is not yet > > converted to private. > > > > load_unaligned_zeropad() can touch memory that is not owned by the > > caller, but just happens to be next after the owned memory. This > > load_unaligned_zeropad() behavior makes it important when the kernel > > asks VMM to convert a GPA from shared to private or back. The kernel > > must never have a page mapped into direct mapping (and aliases) as > > private when the GPA is already converted to shared or when the GPA is > > not yet converted to private. > > Heh, that must be really important info to have it in the changelog twice! > > I'll fix it up when I apply it. Ouch. Please fix the comment in the code too. -- Kiryl Shutsemau / Kirill A. Shutemov
On 6/6/23 11:37, Kirill A. Shutemov wrote: >> Heh, that must be really important info to have it in the changelog twice! >> >> I'll fix it up when I apply it. > Ouch. Please fix the comment in the code too. I couldn't help myself and rewrote the changelog and comment. Please let me know if it looks OK: > https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git/log/?h=x86/tdx
On Tue, Jun 06, 2023 at 04:29:15PM -0700, Dave Hansen wrote: > On 6/6/23 11:37, Kirill A. Shutemov wrote: > >> Heh, that must be really important info to have it in the changelog twice! > >> > >> I'll fix it up when I apply it. > > Ouch. Please fix the comment in the code too. > > I couldn't help myself and rewrote the changelog and comment. Please > let me know if it looks OK: > > > https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git/log/?h=x86/tdx Looks good. Thank you. -- Kiryl Shutsemau / Kirill A. Shutemov
© 2016 - 2026 Red Hat, Inc.