Skip to contentSkip to conversation

hv: Use only one txf buffer per channel and kmalloc on initialize

PATCH In review Hank Janssen 4 participants 6 messages Dec 13, 2010

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 hv: Use only one txf buffer per channel and kmalloc on initialize

Correct issue with not checking kmalloc return value. This fix now only uses one receive buffer for all hv_utils channels, and will do only one kmalloc on init and will return with a -ENOMEM if kmalloc fails on initialize.

And properly clean up memory on failure.

Thanks to Evgeniy Polyakov <[email protected]> for pointing this out. And thanks to Jesper Juhl <[email protected]> and Ky Srinivasan <[email protected]> for suggesting a better implementation of my original patch.

Cc Evgeniy Polyakov <[email protected]>
Cc Jesper Juhl <[email protected]>
Cc Ky Srinivasan <[email protected]>
---
 drivers/staging/hv/hv_utils.c |   84 +++++++++++++++++++++-------------------
 1 files changed, 44 insertions(+), 40 deletions(-)
drivers/staging/hv/hv_utils.c +47 -40
index 53e1e29..e0ecc23 100644
@@ -38,12 +38,14 @@
 #include "vmbus_api.h"
 #include "utils.h"
 
+static u8 *shut_txf_buf;
+static u8 *time_txf_buf;
+static u8 *hbeat_txf_buf;
 
 static void shutdown_onchannelcallback(void *context)
 {
 	struct vmbus_channel *channel = context;
-	u8 *buf;
-	u32 buflen, recvlen;
+	u32 recvlen;
 	u64 requestid;
 	u8  execute_shutdown = false;
 
@@ -52,24 +54,23 @@ static void shutdown_onchannelcallback(void *context)
 	struct icmsg_hdr *icmsghdrp;
 	struct icmsg_negotiate *negop = NULL;
 
-	buflen = PAGE_SIZE;
-	buf = kmalloc(buflen, GFP_ATOMIC);
-
-	vmbus_recvpacket(channel, buf, buflen, &recvlen, &requestid);
+	vmbus_recvpacket(channel, shut_txf_buf,
+			 PAGE_SIZE, &recvlen, &requestid);
 
 	if (recvlen > 0) {
 		DPRINT_DBG(VMBUS, "shutdown packet: len=%d, requestid=%lld",
 			   recvlen, requestid);
 
-		icmsghdrp = (struct icmsg_hdr *)&buf[
+		icmsghdrp = (struct icmsg_hdr *)&shut_txf_buf[
 			sizeof(struct vmbuspipe_hdr)];
 
 		if (icmsghdrp->icmsgtype == ICMSGTYPE_NEGOTIATE) {
-			prep_negotiate_resp(icmsghdrp, negop, buf);
+			prep_negotiate_resp(icmsghdrp, negop, shut_txf_buf);
 		} else {
-			shutdown_msg = (struct shutdown_msg_data *)&buf[
-				sizeof(struct vmbuspipe_hdr) +
-				sizeof(struct icmsg_hdr)];
+			shutdown_msg =
+				(struct shutdown_msg_data *)&shut_txf_buf[
+					sizeof(struct vmbuspipe_hdr) +
+					sizeof(struct icmsg_hdr)];
 
 			switch (shutdown_msg->flags) {
 			case 0:
@@ -93,13 +94,11 @@ static void shutdown_onchannelcallback(void *context)
 		icmsghdrp->icflags = ICMSGHDRFLAG_TRANSACTION
 			| ICMSGHDRFLAG_RESPONSE;
 
-		vmbus_sendpacket(channel, buf,
+		vmbus_sendpacket(channel, shut_txf_buf,
 				       recvlen, requestid,
 				       VmbusPacketTypeDataInBand, 0);
 	}
 
-	kfree(buf);
-
 	if (execute_shutdown == true)
 		orderly_poweroff(false);
 }
@@ -150,28 +149,25 @@ static inline void adj_guesttime(u64 hosttime, u8 flags)
 static void timesync_onchannelcallback(void *context)
 {
 	struct vmbus_channel *channel = context;
-	u8 *buf;
-	u32 buflen, recvlen;
+	u32 recvlen;
 	u64 requestid;
 	struct icmsg_hdr *icmsghdrp;
 	struct ictimesync_data *timedatap;
 
-	buflen = PAGE_SIZE;
-	buf = kmalloc(buflen, GFP_ATOMIC);
-
-	vmbus_recvpacket(channel, buf, buflen, &recvlen, &requestid);
+	vmbus_recvpacket(channel, time_txf_buf,
+			 PAGE_SIZE, &recvlen, &requestid);
 
 	if (recvlen > 0) {
 		DPRINT_DBG(VMBUS, "timesync packet: recvlen=%d, requestid=%lld",
 			recvlen, requestid);
 
-		icmsghdrp = (struct icmsg_hdr *)&buf[
+		icmsghdrp = (struct icmsg_hdr *)&time_txf_buf[
 				sizeof(struct vmbuspipe_hdr)];
 
 		if (icmsghdrp->icmsgtype == ICMSGTYPE_NEGOTIATE) {
-			prep_negotiate_resp(icmsghdrp, NULL, buf);
+			prep_negotiate_resp(icmsghdrp, NULL, time_txf_buf);
 		} else {
-			timedatap = (struct ictimesync_data *)&buf[
+			timedatap = (struct ictimesync_data *)&time_txf_buf[
 				sizeof(struct vmbuspipe_hdr) +
 				sizeof(struct icmsg_hdr)];
 			adj_guesttime(timedatap->parenttime, timedatap->flags);
@@ -180,12 +176,10 @@ static void timesync_onchannelcallback(void *context)
 		icmsghdrp->icflags = ICMSGHDRFLAG_TRANSACTION
 			| ICMSGHDRFLAG_RESPONSE;
 
-		vmbus_sendpacket(channel, buf,
+		vmbus_sendpacket(channel, time_txf_buf,
 				recvlen, requestid,
 				VmbusPacketTypeDataInBand, 0);
 	}
-
-	kfree(buf);
 }
 
 /*
@@ -196,30 +190,28 @@ static void timesync_onchannelcallback(void *context)
 static void heartbeat_onchannelcallback(void *context)
 {
 	struct vmbus_channel *channel = context;
-	u8 *buf;
-	u32 buflen, recvlen;
+	u32 recvlen;
 	u64 requestid;
 	struct icmsg_hdr *icmsghdrp;
 	struct heartbeat_msg_data *heartbeat_msg;
 
-	buflen = PAGE_SIZE;
-	buf = kmalloc(buflen, GFP_ATOMIC);
-
-	vmbus_recvpacket(channel, buf, buflen, &recvlen, &requestid);
+	vmbus_recvpacket(channel, hbeat_txf_buf,
+			 PAGE_SIZE, &recvlen, &requestid);
 
 	if (recvlen > 0) {
 		DPRINT_DBG(VMBUS, "heartbeat packet: len=%d, requestid=%lld",
 			   recvlen, requestid);
 
-		icmsghdrp = (struct icmsg_hdr *)&buf[
+		icmsghdrp = (struct icmsg_hdr *)&hbeat_txf_buf[
 				sizeof(struct vmbuspipe_hdr)];
 
 		if (icmsghdrp->icmsgtype == ICMSGTYPE_NEGOTIATE) {
-			prep_negotiate_resp(icmsghdrp, NULL, buf);
+			prep_negotiate_resp(icmsghdrp, NULL, hbeat_txf_buf);
 		} else {
-			heartbeat_msg = (struct heartbeat_msg_data *)&buf[
-				sizeof(struct vmbuspipe_hdr) +
-				sizeof(struct icmsg_hdr)];
+			heartbeat_msg =
+				(struct heartbeat_msg_data *)&hbeat_txf_buf[
+					sizeof(struct vmbuspipe_hdr) +
+					sizeof(struct icmsg_hdr)];
 
 			DPRINT_DBG(VMBUS, "heartbeat seq = %lld",
 				   heartbeat_msg->seq_num);
@@ -230,12 +222,10 @@ static void heartbeat_onchannelcallback(void *context)
 		icmsghdrp->icflags = ICMSGHDRFLAG_TRANSACTION
 			| ICMSGHDRFLAG_RESPONSE;
 
-		vmbus_sendpacket(channel, buf,
+		vmbus_sendpacket(channel, hbeat_txf_buf,
 				       recvlen, requestid,
 				       VmbusPacketTypeDataInBand, 0);
 	}
-
-	kfree(buf);
 }
 
 static const struct pci_device_id __initconst
@@ -268,6 +258,19 @@ static int __init init_hyperv_utils(void)
 	if (!dmi_check_system(hv_utils_dmi_table))
 		return -ENODEV;
 
+	shut_txf_buf = kmalloc(PAGE_SIZE, GFP_ATOMIC);
+	time_txf_buf = kmalloc(PAGE_SIZE, GFP_ATOMIC);
+	hbeat_txf_buf = kmalloc(PAGE_SIZE, GFP_ATOMIC);
+
+	if (!shut_txf_buf || !time_txf_buf || !hbeat_txf_buf) {
+		printk(KERN_INFO
+		       "Unable to allocate memory for receive buffer\n");
+		kfree(shut_txf_buf);
+		kfree(time_txf_buf);
+		kfree(hbeat_txf_buf);
+		return -ENOMEM;
+	}
+
 	hv_cb_utils[HV_SHUTDOWN_MSG].channel->onchannel_callback =
 		&shutdown_onchannelcallback;
 	hv_cb_utils[HV_SHUTDOWN_MSG].callback = &shutdown_onchannelcallback;
@@ -298,6 +298,10 @@ static void exit_hyperv_utils(void)
 	hv_cb_utils[HV_HEARTBEAT_MSG].channel->onchannel_callback =
 		&chn_cb_negotiate;
 	hv_cb_utils[HV_HEARTBEAT_MSG].callback = &chn_cb_negotiate;
+
+	kfree(shut_txf_buf);
+	kfree(time_txf_buf);
+	kfree(hbeat_txf_buf);
 }
 
 module_init(init_hyperv_utils);
Signature · 1 line
1.6.0.2
Reviewed-by on v1 1/1
#1 lore raw
Quoted 12 lines · Hank Janssen

Correct issue with not checking kmalloc return value. This fix now only uses one receive buffer for all hv_utils channels, and will do only one kmalloc on init and will return with a -ENOMEM if kmalloc fails on initialize.

And properly clean up memory on failure.

Thanks to Evgeniy Polyakov <[email protected]> for pointing this out. And thanks to Jesper Juhl <[email protected]> and Ky Srinivasan <[email protected]> for suggesting a better implementation of my original patch.

Quoted 6 lines
Cc Evgeniy Polyakov <[email protected]>
Cc Jesper Juhl <[email protected]>
Cc Ky Srinivasan <[email protected]>

I can't spot any problems with these changes now, so feel free to add

if you like.

Signature · 3 lines
Jesper Juhl <[email protected]>            http://www.chaosbits.net/
Don't top-post http://www.catb.org/~esr/jargon/html/T/top-post.html
Plain text mails only, please.
#2 lore raw
Quoted 7 lines

From: Jesper Juhl [mailto:[email protected]]
Sent: Monday, December 13, 2010 1:22 PM

I can't spot any problems with these changes now, so feel free to add

Thank you for your help Jesper.

Greg, do you want me to resubmit with the Reviewed by line? Or do you Add it once you queue it up for linux-next?

Thanks,

Hank.

#3 lore raw
Quoted 4 lines · Hank Janssen
Quoted 4 lines

From: Jesper Juhl [mailto:[email protected]]
Sent: Monday, December 13, 2010 1:22 PM

Quoted 8 lines
Quoted 3 lines

I can't spot any problems with these changes now, so feel free to add

Thank you for your help Jesper.

Greg, do you want me to resubmit with the Reviewed by line? Or do you Add it once you queue it up for linux-next?

I add it when I apply it.

thanks,

greg k-h

#4 lore raw
In reply to Hank Janssen

drivers/staging/hv/hv_utils.c line 43

Quoted 3 lines · Hank Janssen
drivers/staging/hv/hv_utils.c +3 -0
+static u8 *shut_txf_buf;
+static u8 *time_txf_buf;
+static u8 *hbeat_txf_buf;

Those are accessed without any kind of synchronization. I do not know exact context, but are you sure it is single-threaded?

Signature · 1 line
	Evgeniy Polyakov
#5 lore raw

-----Original Message-----

Quoted 9 lines

From: Evgeniy Polyakov [mailto:[email protected]]
Sent: Monday, December 13, 2010 2:04 PM

drivers/staging/hv/hv_utils.c line 43

Quoted 3 lines · Hank Janssen
drivers/staging/hv/hv_utils.c +3 -0
+static u8 *shut_txf_buf;
+static u8 *time_txf_buf;
+static u8 *hbeat_txf_buf;

Those are accessed without any kind of synchronization. I do not know exact context, but are you sure it is single-threaded?

The hv_utils driver services are all single threaded. The other drivers are not, but this one is.

Hank.