diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2023-04-25 10:48:08 -0700 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2023-04-25 10:48:08 -0700 |
commit | bc1bb2a49bae915107fd58705edb1e32db92c635 (patch) | |
tree | 2b517448e60c76160452349ea14766ae309dfba0 /init | |
parent | c42b59bfaa0091833b6758be772c54ec7183daa5 (diff) | |
parent | 812b0597fb4043240724e4c7bed7ba1fe15c0e3f (diff) | |
download | lwn-bc1bb2a49bae915107fd58705edb1e32db92c635.tar.gz lwn-bc1bb2a49bae915107fd58705edb1e32db92c635.zip |
Merge tag 'x86_sev_for_v6.4_rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 SEV updates from Borislav Petkov:
- Add the necessary glue so that the kernel can run as a confidential
SEV-SNP vTOM guest on Hyper-V. A vTOM guest basically splits the
address space in two parts: encrypted and unencrypted. The use case
being running unmodified guests on the Hyper-V confidential computing
hypervisor
- Double-buffer messages between the guest and the hardware PSP device
so that no partial buffers are copied back'n'forth and thus potential
message integrity and leak attacks are possible
- Name the return value the sev-guest driver returns when the hw PSP
device hasn't been called, explicitly
- Cleanups
* tag 'x86_sev_for_v6.4_rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/hyperv: Change vTOM handling to use standard coco mechanisms
init: Call mem_encrypt_init() after Hyper-V hypercall init is done
x86/mm: Handle decryption/re-encryption of bss_decrypted consistently
Drivers: hv: Explicitly request decrypted in vmap_pfn() calls
x86/hyperv: Reorder code to facilitate future work
x86/ioremap: Add hypervisor callback for private MMIO mapping in coco VM
x86/sev: Change snp_guest_issue_request()'s fw_err argument
virt/coco/sev-guest: Double-buffer messages
crypto: ccp: Get rid of __sev_platform_init_locked()'s local function pointer
crypto: ccp - Name -1 return value as SEV_RET_NO_FW_CALL
Diffstat (limited to 'init')
-rw-r--r-- | init/main.c | 19 |
1 files changed, 11 insertions, 8 deletions
diff --git a/init/main.c b/init/main.c index 123f50eb362b..c62f0c8811d7 100644 --- a/init/main.c +++ b/init/main.c @@ -1092,14 +1092,6 @@ asmlinkage __visible void __init __no_sanitize_address start_kernel(void) */ locking_selftest(); - /* - * This needs to be called before any devices perform DMA - * operations that might use the SWIOTLB bounce buffers. It will - * mark the bounce buffers as decrypted so that their usage will - * not cause "plain-text" data to be decrypted when accessed. - */ - mem_encrypt_init(); - #ifdef CONFIG_BLK_DEV_INITRD if (initrd_start && !initrd_below_start_ok && page_to_pfn(virt_to_page((void *)initrd_start)) < min_low_pfn) { @@ -1116,6 +1108,17 @@ asmlinkage __visible void __init __no_sanitize_address start_kernel(void) late_time_init(); sched_clock_init(); calibrate_delay(); + + /* + * This needs to be called before any devices perform DMA + * operations that might use the SWIOTLB bounce buffers. It will + * mark the bounce buffers as decrypted so that their usage will + * not cause "plain-text" data to be decrypted when accessed. It + * must be called after late_time_init() so that Hyper-V x86/x64 + * hypercalls work when the SWIOTLB bounce buffers are decrypted. + */ + mem_encrypt_init(); + pid_idr_init(); anon_vma_init(); #ifdef CONFIG_X86 |