mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: yangge1116@126.com
To: pbonzini@redhat.com
Cc: kvm@vger.kernel.org, linux-kernel@vger.kernel.org,
	seanjc@google.com, 21cnbao@gmail.com, david@redhat.com,
	baolin.wang@linux.alibaba.com, liuzixing@hygon.cn,
	yangge <yangge1116@126.com>
Subject: [PATCH V3] KVM: SEV: fix wrong pinning of pages
Date: Sat, 11 Jan 2025 14:15:22 +0800	[thread overview]
Message-ID: <1736576122-9818-1-git-send-email-yangge1116@126.com> (raw)

From: yangge <yangge1116@126.com>

In the sev_mem_enc_register_region() function, we need to call
sev_pin_memory() to pin memory for the long term. However, when
calling sev_pin_memory(), the FOLL_LONGTERM flag is not passed, causing
the allocated pages not to be migrated out of MIGRATE_CMA/ZONE_MOVABLE,
violating these mechanisms to avoid fragmentation with unmovable pages,
for example making CMA allocations fail.

To address the aforementioned problem, we should add the FOLL_LONGTERM
flag when calling sev_pin_memory() within the sev_mem_enc_register_region()
function.

Signed-off-by: yangge <yangge1116@126.com>
---

V3:
- the fix only needed for sev_mem_enc_register_region()

V2:
- update code and commit message suggested by David

 arch/x86/kvm/svm/sev.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)

diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c
index 943bd07..04a125c 100644
--- a/arch/x86/kvm/svm/sev.c
+++ b/arch/x86/kvm/svm/sev.c
@@ -622,7 +622,7 @@ static int sev_launch_start(struct kvm *kvm, struct kvm_sev_cmd *argp)
 
 static struct page **sev_pin_memory(struct kvm *kvm, unsigned long uaddr,
 				    unsigned long ulen, unsigned long *n,
-				    int write)
+				    int flags)
 {
 	struct kvm_sev_info *sev = &to_kvm_svm(kvm)->sev_info;
 	unsigned long npages, size;
@@ -663,7 +663,7 @@ static struct page **sev_pin_memory(struct kvm *kvm, unsigned long uaddr,
 		return ERR_PTR(-ENOMEM);
 
 	/* Pin the user virtual address. */
-	npinned = pin_user_pages_fast(uaddr, npages, write ? FOLL_WRITE : 0, pages);
+	npinned = pin_user_pages_fast(uaddr, npages, flags, pages);
 	if (npinned != npages) {
 		pr_err("SEV: Failure locking %lu pages.\n", npages);
 		ret = -ENOMEM;
@@ -751,7 +751,7 @@ static int sev_launch_update_data(struct kvm *kvm, struct kvm_sev_cmd *argp)
 	vaddr_end = vaddr + size;
 
 	/* Lock the user memory. */
-	inpages = sev_pin_memory(kvm, vaddr, size, &npages, 1);
+	inpages = sev_pin_memory(kvm, vaddr, size, &npages, FOLL_WRITE);
 	if (IS_ERR(inpages))
 		return PTR_ERR(inpages);
 
@@ -1250,7 +1250,8 @@ static int sev_dbg_crypt(struct kvm *kvm, struct kvm_sev_cmd *argp, bool dec)
 		if (IS_ERR(src_p))
 			return PTR_ERR(src_p);
 
-		dst_p = sev_pin_memory(kvm, dst_vaddr & PAGE_MASK, PAGE_SIZE, &n, 1);
+		dst_p = sev_pin_memory(kvm, dst_vaddr & PAGE_MASK, PAGE_SIZE, &n,
+					FOLL_WRITE);
 		if (IS_ERR(dst_p)) {
 			sev_unpin_memory(kvm, src_p, n);
 			return PTR_ERR(dst_p);
@@ -1316,7 +1317,7 @@ static int sev_launch_secret(struct kvm *kvm, struct kvm_sev_cmd *argp)
 	if (copy_from_user(&params, u64_to_user_ptr(argp->data), sizeof(params)))
 		return -EFAULT;
 
-	pages = sev_pin_memory(kvm, params.guest_uaddr, params.guest_len, &n, 1);
+	pages = sev_pin_memory(kvm, params.guest_uaddr, params.guest_len, &n, FOLL_WRITE);
 	if (IS_ERR(pages))
 		return PTR_ERR(pages);
 
@@ -1798,7 +1799,7 @@ static int sev_receive_update_data(struct kvm *kvm, struct kvm_sev_cmd *argp)
 
 	/* Pin guest memory */
 	guest_page = sev_pin_memory(kvm, params.guest_uaddr & PAGE_MASK,
-				    PAGE_SIZE, &n, 1);
+				    PAGE_SIZE, &n, FOLL_WRITE);
 	if (IS_ERR(guest_page)) {
 		ret = PTR_ERR(guest_page);
 		goto e_free_trans;
@@ -2696,7 +2697,8 @@ int sev_mem_enc_register_region(struct kvm *kvm,
 		return -ENOMEM;
 
 	mutex_lock(&kvm->lock);
-	region->pages = sev_pin_memory(kvm, range->addr, range->size, &region->npages, 1);
+	region->pages = sev_pin_memory(kvm, range->addr, range->size, &region->npages,
+				FOLL_WRITE | FOLL_LONGTERM);
 	if (IS_ERR(region->pages)) {
 		ret = PTR_ERR(region->pages);
 		mutex_unlock(&kvm->lock);
-- 
2.7.4


             reply	other threads:[~2025-01-11  6:16 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-01-11  6:15 yangge1116 [this message]
2025-02-10 21:37 ` Tom Lendacky

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=1736576122-9818-1-git-send-email-yangge1116@126.com \
    --to=yangge1116@126.com \
    --cc=21cnbao@gmail.com \
    --cc=baolin.wang@linux.alibaba.com \
    --cc=david@redhat.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=liuzixing@hygon.cn \
    --cc=pbonzini@redhat.com \
    --cc=seanjc@google.com \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®