Linux-HyperV Archive on lore.kernel.org
 help / color / Atom feed
* [PATCH net-next v2] hv_sock: use HV_HYP_PAGE_SIZE for Hyper-V communication
@ 2019-10-13  0:30 Michael Kelley
  2019-10-16  0:27 ` David Miller
  0 siblings, 1 reply; 2+ messages in thread
From: Michael Kelley @ 2019-10-13  0:30 UTC (permalink / raw)
  To: KY Srinivasan, Haiyang Zhang, Stephen Hemminger, sashal, davem,
	linux-hyperv, netdev, linux-kernel
  Cc: himadrispandya

From: Himadri Pandya <himadrispandya@gmail.com>

Current code assumes PAGE_SIZE (the guest page size) is equal
to the page size used to communicate with Hyper-V (which is
always 4K). While this assumption is true on x86, it may not
be true for Hyper-V on other architectures. For example,
Linux on ARM64 may have PAGE_SIZE of 16K or 64K. A new symbol,
HV_HYP_PAGE_SIZE, has been previously introduced to use when
the Hyper-V page size is intended instead of the guest page size.

Make this code work on non-x86 architectures by using the new
HV_HYP_PAGE_SIZE symbol instead of PAGE_SIZE, where appropriate.
Also replace the now redundant PAGE_SIZE_4K with HV_HYP_PAGE_SIZE.
The change has no effect on x86, but lays the groundwork to run
on ARM64 and others.

Signed-off-by: Himadri Pandya <himadrispandya@gmail.com>
Reviewed-by: Michael Kelley <mikelley@microsoft.com>
---

Changes in v2:
* Revised commit message and subject [Jakub Kicinski]

---
 net/vmw_vsock/hyperv_transport.c | 22 ++++++++++++----------
 1 file changed, 12 insertions(+), 10 deletions(-)

diff --git a/net/vmw_vsock/hyperv_transport.c b/net/vmw_vsock/hyperv_transport.c
index 261521d..d2929ea 100644
--- a/net/vmw_vsock/hyperv_transport.c
+++ b/net/vmw_vsock/hyperv_transport.c
@@ -13,15 +13,16 @@
 #include <linux/hyperv.h>
 #include <net/sock.h>
 #include <net/af_vsock.h>
+#include <asm/hyperv-tlfs.h>
 
 /* Older (VMBUS version 'VERSION_WIN10' or before) Windows hosts have some
- * stricter requirements on the hv_sock ring buffer size of six 4K pages. Newer
- * hosts don't have this limitation; but, keep the defaults the same for compat.
+ * stricter requirements on the hv_sock ring buffer size of six 4K pages.
+ * hyperv-tlfs defines HV_HYP_PAGE_SIZE as 4K. Newer hosts don't have this
+ * limitation; but, keep the defaults the same for compat.
  */
-#define PAGE_SIZE_4K		4096
-#define RINGBUFFER_HVS_RCV_SIZE (PAGE_SIZE_4K * 6)
-#define RINGBUFFER_HVS_SND_SIZE (PAGE_SIZE_4K * 6)
-#define RINGBUFFER_HVS_MAX_SIZE (PAGE_SIZE_4K * 64)
+#define RINGBUFFER_HVS_RCV_SIZE (HV_HYP_PAGE_SIZE * 6)
+#define RINGBUFFER_HVS_SND_SIZE (HV_HYP_PAGE_SIZE * 6)
+#define RINGBUFFER_HVS_MAX_SIZE (HV_HYP_PAGE_SIZE * 64)
 
 /* The MTU is 16KB per the host side's design */
 #define HVS_MTU_SIZE		(1024 * 16)
@@ -54,7 +55,8 @@ struct hvs_recv_buf {
  * ringbuffer APIs that allow us to directly copy data from userspace buffer
  * to VMBus ringbuffer.
  */
-#define HVS_SEND_BUF_SIZE (PAGE_SIZE_4K - sizeof(struct vmpipe_proto_header))
+#define HVS_SEND_BUF_SIZE \
+		(HV_HYP_PAGE_SIZE - sizeof(struct vmpipe_proto_header))
 
 struct hvs_send_buf {
 	/* The header before the payload data */
@@ -393,10 +395,10 @@ static void hvs_open_connection(struct vmbus_channel *chan)
 	} else {
 		sndbuf = max_t(int, sk->sk_sndbuf, RINGBUFFER_HVS_SND_SIZE);
 		sndbuf = min_t(int, sndbuf, RINGBUFFER_HVS_MAX_SIZE);
-		sndbuf = ALIGN(sndbuf, PAGE_SIZE);
+		sndbuf = ALIGN(sndbuf, HV_HYP_PAGE_SIZE);
 		rcvbuf = max_t(int, sk->sk_rcvbuf, RINGBUFFER_HVS_RCV_SIZE);
 		rcvbuf = min_t(int, rcvbuf, RINGBUFFER_HVS_MAX_SIZE);
-		rcvbuf = ALIGN(rcvbuf, PAGE_SIZE);
+		rcvbuf = ALIGN(rcvbuf, HV_HYP_PAGE_SIZE);
 	}
 
 	ret = vmbus_open(chan, sndbuf, rcvbuf, NULL, 0, hvs_channel_cb,
@@ -670,7 +672,7 @@ static ssize_t hvs_stream_enqueue(struct vsock_sock *vsk, struct msghdr *msg,
 	ssize_t ret = 0;
 	ssize_t bytes_written = 0;
 
-	BUILD_BUG_ON(sizeof(*send_buf) != PAGE_SIZE_4K);
+	BUILD_BUG_ON(sizeof(*send_buf) != HV_HYP_PAGE_SIZE);
 
 	send_buf = kmalloc(sizeof(*send_buf), GFP_KERNEL);
 	if (!send_buf)
-- 
1.8.3.1


^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [PATCH net-next v2] hv_sock: use HV_HYP_PAGE_SIZE for Hyper-V communication
  2019-10-13  0:30 [PATCH net-next v2] hv_sock: use HV_HYP_PAGE_SIZE for Hyper-V communication Michael Kelley
@ 2019-10-16  0:27 ` David Miller
  0 siblings, 0 replies; 2+ messages in thread
From: David Miller @ 2019-10-16  0:27 UTC (permalink / raw)
  To: mikelley
  Cc: kys, haiyangz, sthemmin, sashal, linux-hyperv, netdev,
	linux-kernel, himadrispandya

From: Michael Kelley <mikelley@microsoft.com>
Date: Sun, 13 Oct 2019 00:30:21 +0000

> From: Himadri Pandya <himadrispandya@gmail.com>
> 
> Current code assumes PAGE_SIZE (the guest page size) is equal
> to the page size used to communicate with Hyper-V (which is
> always 4K). While this assumption is true on x86, it may not
> be true for Hyper-V on other architectures. For example,
> Linux on ARM64 may have PAGE_SIZE of 16K or 64K. A new symbol,
> HV_HYP_PAGE_SIZE, has been previously introduced to use when
> the Hyper-V page size is intended instead of the guest page size.
> 
> Make this code work on non-x86 architectures by using the new
> HV_HYP_PAGE_SIZE symbol instead of PAGE_SIZE, where appropriate.
> Also replace the now redundant PAGE_SIZE_4K with HV_HYP_PAGE_SIZE.
> The change has no effect on x86, but lays the groundwork to run
> on ARM64 and others.
> 
> Signed-off-by: Himadri Pandya <himadrispandya@gmail.com>
> Reviewed-by: Michael Kelley <mikelley@microsoft.com>
> ---
> 
> Changes in v2:
> * Revised commit message and subject [Jakub Kicinski]

Applied, thank you.

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, back to index

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-10-13  0:30 [PATCH net-next v2] hv_sock: use HV_HYP_PAGE_SIZE for Hyper-V communication Michael Kelley
2019-10-16  0:27 ` David Miller

Linux-HyperV Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/linux-hyperv/0 linux-hyperv/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 linux-hyperv linux-hyperv/ https://lore.kernel.org/linux-hyperv \
		linux-hyperv@vger.kernel.org
	public-inbox-index linux-hyperv

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.linux-hyperv


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git