RE: [PATCH v4 2/6] x86/tdx: Support vmalloc() for tdx_enc_status_changed()

From: Michael Kelley (LINUX)
Date: Tue Apr 11 2023 - 12:28:29 EST


From: Dexuan Cui <decui@xxxxxxxxxxxxx> Sent: Saturday, April 8, 2023 1:48 PM
>
> When a TDX guest runs on Hyper-V, the hv_netvsc driver's netvsc_init_buf()
> allocates buffers using vzalloc(), and needs to share the buffers with the
> host OS by calling set_memory_decrypted(), which is not working for
> vmalloc() yet. Add the support by handling the pages one by one.
>
> Co-developed-by: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>
> Signed-off-by: Dexuan Cui <decui@xxxxxxxxxxxxx>
> ---
> arch/x86/coco/tdx/tdx.c | 76 ++++++++++++++++++++++++++++-------------
> 1 file changed, 52 insertions(+), 24 deletions(-)
>
> Changes in v2:
> Changed tdx_enc_status_changed() in place.
>
> Hi, Dave, I checked the huge vmalloc mapping code, but still don't know
> how to get the underlying huge page info (if huge page is in use) and
> try to use PG_LEVEL_2M/1G in try_accept_page() for vmalloc: I checked
> is_vm_area_hugepages() and __vfree() -> __vunmap(), and I think the
> underlying page allocation info is internal to the mm code, and there
> is no mm API to for me get the info in tdx_enc_status_changed().
>
> Changes in v3:
> No change since v2.
>
> Changes in v4:
> Added Kirill's Co-developed-by since Kirill helped to improve the
> code by adding tdx_enc_status_changed_phys().
>
> Thanks Kirill for the clarification on load_unaligned_zeropad()!
>
> The vzalloc() usage in drivers/net/hyperv/netvsc.c: netvsc_init_buf()
> remains the same. It may not worth it to "allocate a vmalloc region,
> allocate pages manually", because we have to consider the worst case
> where the system is sufferiing from severe memory fragmentation and
> we can only allocate multiple single pages. We may not want to
> complicate the code in netvsc_init_buf(). We'll support NIC SR-IOV
> for TDX VMs on Hyper-V, so the netvsc send/recv buffers won't be
> used when the VF NIC is up.
>
> diff --git a/arch/x86/coco/tdx/tdx.c b/arch/x86/coco/tdx/tdx.c
> index 5574c91541a2d..731be50b3d093 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/mm.h>
> #include <asm/coco.h>
> #include <asm/tdx.h>
> #include <asm/vmx.h>
> @@ -789,6 +790,34 @@ static bool try_accept_one(phys_addr_t *start, unsigned long
> len,
> return true;
> }
>
> +static bool try_accept_page(phys_addr_t start, phys_addr_t end)
> +{
> + /*
> + * For shared->private conversion, accept the page using
> + * TDX_ACCEPT_PAGE TDX module call.
> + */
> + while (start < end) {
> + unsigned long len = end - start;
> +
> + /*
> + * Try larger accepts first. It gives chance to VMM to keep
> + * 1G/2M SEPT entries where possible and speeds up process by
> + * cutting number of hypercalls (if successful).
> + */
> +
> + if (try_accept_one(&start, len, PG_LEVEL_1G))
> + continue;
> +
> + if (try_accept_one(&start, len, PG_LEVEL_2M))
> + continue;
> +
> + if (!try_accept_one(&start, len, PG_LEVEL_4K))
> + return false;
> + }
> +
> + return true;
> +}
> +
> /*
> * Notify the VMM about page mapping conversion. More info about ABI
> * can be found in TDX Guest-Host-Communication Interface (GHCI),
> @@ -838,6 +867,19 @@ static bool tdx_map_gpa(phys_addr_t start, phys_addr_t
> end, bool enc)
> return !ret;
> }
>
> +static bool tdx_enc_status_changed_phys(phys_addr_t start, phys_addr_t end,
> + bool enc)
> +{
> + if (!tdx_map_gpa(start, end, enc))
> + return false;
> +
> + /* private->shared conversion requires only MapGPA call */
> + if (!enc)
> + return true;
> +
> + return try_accept_page(start, end);
> +}
> +
> /*
> * Inform the VMM of the guest's intent for this physical page: shared with
> * the VMM or private to the guest. The VMM is expected to change its mapping
> @@ -845,37 +887,23 @@ static bool tdx_map_gpa(phys_addr_t start, phys_addr_t
> end, bool enc)
> */
> static bool tdx_enc_status_changed(unsigned long vaddr, int numpages, bool enc)
> {
> - phys_addr_t start = __pa(vaddr);
> - phys_addr_t end = __pa(vaddr + numpages * PAGE_SIZE);
> + unsigned long start = vaddr;
> + unsigned long end = start + numpages * PAGE_SIZE;
>
> - if (!tdx_map_gpa(start, end, enc))
> + if (offset_in_page(start) != 0)
> return false;
>
> - /* private->shared conversion requires only MapGPA call */
> - if (!enc)
> - return true;
> + if (!is_vmalloc_addr((void *)start))
> + return tdx_enc_status_changed_phys(__pa(start), __pa(end), enc);
>
> - /*
> - * For shared->private conversion, accept the page using
> - * TDX_ACCEPT_PAGE TDX module call.
> - */
> while (start < end) {
> - unsigned long len = end - start;
> + phys_addr_t start_pa = slow_virt_to_phys((void *)start);

I've usually seen "page_to_phys(vmalloc_to_page(start))" for getting
the physical address corresponding to a vmalloc() address. For example,
see virt_to_hvpfn(), kvm_kaddr_to_phys(), nx842_get_pa(),
rproc_va_to_pa(), etc.

Does anyone know if there is a reason to use one vs. the other?
slow_virt_to_phys() is x86-only, but that's not a problem here.

Otherwise,
Reviewed-by: Michael Kelley <mikelley@xxxxxxxxxxxxx>

> + phys_addr_t end_pa = start_pa + PAGE_SIZE;
>
> - /*
> - * Try larger accepts first. It gives chance to VMM to keep
> - * 1G/2M SEPT entries where possible and speeds up process by
> - * cutting number of hypercalls (if successful).
> - */
> -
> - if (try_accept_one(&start, len, PG_LEVEL_1G))
> - continue;
> -
> - if (try_accept_one(&start, len, PG_LEVEL_2M))
> - continue;
> -
> - if (!try_accept_one(&start, len, PG_LEVEL_4K))
> + if (!tdx_enc_status_changed_phys(start_pa, end_pa, enc))
> return false;
> +
> + start += PAGE_SIZE;
> }
>
> return true;
> --
> 2.25.1