Merge from-linus to-akpm

arch/x86_64/kernel/smpboot.c

@@ -334,7 +334,7 @@ static void __cpuinit tsc_sync_wait(void)
 {
 	if (notscsync || !cpu_has_tsc)
 		return;
-	sync_tsc(boot_cpu_id);
+	sync_tsc(0);
 }
 
 static __init int notscsync_setup(char *s)

drivers/acpi/motherboard.c

@@ -42,7 +42,7 @@ ACPI_MODULE_NAME("acpi_motherboard")
  */
 #define IS_RESERVED_ADDR(base, len) \
 	(((len) > 0) && ((base) > 0) && ((base) + (len) < IO_SPACE_LIMIT) \
-	&& ((base) + (len) > 0x1000))
+	&& ((base) + (len) > PCIBIOS_MIN_IO))
 /*
  * Clearing the flag (IORESOURCE_BUSY) allows drivers to use
  * the io ports if they really know they can use it, while

drivers/char/mem.c

@@ -261,7 +261,11 @@ static int mmap_mem(struct file * file, struct vm_area_struct * vma)
 
 static int mmap_kmem(struct file * file, struct vm_area_struct * vma)
 {
-        unsigned long long val;
+	unsigned long pfn;
+
+	/* Turn a kernel-virtual address into a physical page frame */
+	pfn = __pa((u64)vma->vm_pgoff << PAGE_SHIFT) >> PAGE_SHIFT;
+
 	/*
 	 * RED-PEN: on some architectures there is more mapped memory
 	 * than available in mem_map which pfn_valid checks
@@ -269,10 +273,10 @@ static int mmap_kmem(struct file * file, struct vm_area_struct * vma)
 	 *
 	 * RED-PEN: vmalloc is not supported right now.
 	 */
-	if (!pfn_valid(vma->vm_pgoff))
+	if (!pfn_valid(pfn))
 		return -EIO;
-	val = (u64)vma->vm_pgoff << PAGE_SHIFT;
-	vma->vm_pgoff = __pa(val) >> PAGE_SHIFT;
+
+	vma->vm_pgoff = pfn;
 	return mmap_mem(file, vma);
 }
 

drivers/net/e1000/e1000_main.c

@@ -3789,6 +3789,7 @@ e1000_netpoll(struct net_device *netdev)
 	struct e1000_adapter *adapter = netdev_priv(netdev);
 	disable_irq(adapter->pdev->irq);
 	e1000_intr(adapter->pdev->irq, netdev, NULL);
+	e1000_clean_tx_irq(adapter);
 	enable_irq(adapter->pdev->irq);
 }
 #endif

drivers/scsi/dc395x.c

@@ -183,7 +183,7 @@
  * cross a page boundy.
  */
 #define SEGMENTX_LEN	(sizeof(struct SGentry)*DC395x_MAX_SG_LISTENTRY)
-#define VIRTX_LEN	(sizeof(void *) * DC395x_MAX_SG_LISTENTRY)
+
 
 struct SGentry {
 	u32 address;		/* bus! address */
@@ -235,7 +235,6 @@ struct ScsiReqBlk {
 	u8 sg_count;			/* No of HW sg entries for this request */
 	u8 sg_index;			/* Index of HW sg entry for this request */
 	u32 total_xfer_length;		/* Total number of bytes remaining to be transfered */
-	void **virt_map;
 	unsigned char *virt_addr;	/* Virtual address of current transfer position */
 
 	/*
@@ -1022,14 +1021,14 @@ static void build_srb(struct scsi_cmnd *cmd, struct DeviceCtlBlk *dcb,
 			reqlen, cmd->request_buffer, cmd->use_sg,
 			srb->sg_count);
 
+		srb->virt_addr = page_address(sl->page);
 		for (i = 0; i < srb->sg_count; i++) {
-			u32 seglen = (u32)sg_dma_len(sl + i);
-			sgp[i].address = (u32)sg_dma_address(sl + i);
+			u32 busaddr = (u32)sg_dma_address(&sl[i]);
+			u32 seglen = (u32)sl[i].length;
+			sgp[i].address = busaddr;
 			sgp[i].length = seglen;
 			srb->total_xfer_length += seglen;
-			srb->virt_map[i] = kmap(sl[i].page);
 		}
-		srb->virt_addr = srb->virt_map[0];
 		sgp += srb->sg_count - 1;
 
 		/*
@@ -1976,7 +1975,6 @@ static void sg_update_list(struct ScsiReqBlk *srb, u32 left)
 	int segment = cmd->use_sg;
 	u32 xferred = srb->total_xfer_length - left; /* bytes transfered */
 	struct SGentry *psge = srb->segment_x + srb->sg_index;
-	void **virt = srb->virt_map;
 
 	dprintkdbg(DBG_0,
 		"sg_update_list: Transfered %i of %i bytes, %i remain\n",
@@ -2016,16 +2014,16 @@ static void sg_update_list(struct ScsiReqBlk *srb, u32 left)
 
 	/* We have to walk the scatterlist to find it */
 	sg = (struct scatterlist *)cmd->request_buffer;
-	idx = 0;
 	while (segment--) {
 		unsigned long mask =
 		    ~((unsigned long)sg->length - 1) & PAGE_MASK;
 		if ((sg_dma_address(sg) & mask) == (psge->address & mask)) {
-			srb->virt_addr = virt[idx] + (psge->address & ~PAGE_MASK);
+			srb->virt_addr = (page_address(sg->page)
+					   + psge->address -
+					   (psge->address & PAGE_MASK));
 			return;
 		}
 		++sg;
-		++idx;
 	}
 
 	dprintkl(KERN_ERR, "sg_update_list: sg_to_virt failed\n");
@@ -2151,7 +2149,7 @@ static void data_out_phase0(struct AdapterCtlBlk *acb, struct ScsiReqBlk *srb,
 				DC395x_read32(acb, TRM_S1040_DMA_CXCNT));
 		}
 		/*
-		 * calculate all the residue data that not yet transfered
+		 * calculate all the residue data that not yet tranfered
 		 * SCSI transfer counter + left in SCSI FIFO data
 		 *
 		 * .....TRM_S1040_SCSI_COUNTER (24bits)
@@ -3269,7 +3267,6 @@ static void pci_unmap_srb(struct AdapterCtlBlk *acb, struct ScsiReqBlk *srb)
 	struct scsi_cmnd *cmd = srb->cmd;
 	enum dma_data_direction dir = cmd->sc_data_direction;
 	if (cmd->use_sg && dir != PCI_DMA_NONE) {
-		int i;
 		/* unmap DC395x SG list */
 		dprintkdbg(DBG_SG, "pci_unmap_srb: list=%08x(%05x)\n",
 			srb->sg_bus_addr, SEGMENTX_LEN);
@@ -3279,8 +3276,6 @@ static void pci_unmap_srb(struct AdapterCtlBlk *acb, struct ScsiReqBlk *srb)
 		dprintkdbg(DBG_SG, "pci_unmap_srb: segs=%i buffer=%p\n",
 			cmd->use_sg, cmd->request_buffer);
 		/* unmap the sg segments */
-		for (i = 0; i < srb->sg_count; i++)
-			kunmap(virt_to_page(srb->virt_map[i]));
 		pci_unmap_sg(acb->dev,
 			     (struct scatterlist *)cmd->request_buffer,
 			     cmd->use_sg, dir);
@@ -3327,7 +3322,7 @@ static void srb_done(struct AdapterCtlBlk *acb, struct DeviceCtlBlk *dcb,
 
 	if (cmd->use_sg) {
 		struct scatterlist* sg = (struct scatterlist *)cmd->request_buffer;
-		ptr = (struct ScsiInqData *)(srb->virt_map[0] + sg->offset);
+		ptr = (struct ScsiInqData *)(page_address(sg->page) + sg->offset);
 	} else {
 		ptr = (struct ScsiInqData *)(cmd->request_buffer);
 	}
@@ -4262,9 +4257,8 @@ static void adapter_sg_tables_free(struct AdapterCtlBlk *acb)
 	const unsigned srbs_per_page = PAGE_SIZE/SEGMENTX_LEN;
 
 	for (i = 0; i < DC395x_MAX_SRB_CNT; i += srbs_per_page)
+		if (acb->srb_array[i].segment_x)
 			kfree(acb->srb_array[i].segment_x);
-
-	vfree(acb->srb_array[0].virt_map);
 }
 
 
@@ -4280,12 +4274,9 @@ static int __devinit adapter_sg_tables_alloc(struct AdapterCtlBlk *acb)
 	int srb_idx = 0;
 	unsigned i = 0;
 	struct SGentry *ptr;
-	void **virt_array;
 
-	for (i = 0; i < DC395x_MAX_SRB_CNT; i++) {
+	for (i = 0; i < DC395x_MAX_SRB_CNT; i++)
 		acb->srb_array[i].segment_x = NULL;
-		acb->srb_array[i].virt_map = NULL;
-	}
 
 	dprintkdbg(DBG_1, "Allocate %i pages for SG tables\n", pages);
 	while (pages--) {
@@ -4306,19 +4297,6 @@ static int __devinit adapter_sg_tables_alloc(struct AdapterCtlBlk *acb)
 		    ptr + (i * DC395x_MAX_SG_LISTENTRY);
 	else
 		dprintkl(KERN_DEBUG, "No space for tmsrb SG table reserved?!\n");
-
-	virt_array = vmalloc((DC395x_MAX_SRB_CNT + 1) * DC395x_MAX_SG_LISTENTRY * sizeof(void*));
-
-	if (!virt_array) {
-		adapter_sg_tables_free(acb);
-		return 1;
-	}
-
-	for (i = 0; i < DC395x_MAX_SRB_CNT + 1; i++) {
-		acb->srb_array[i].virt_map = virt_array;
-		virt_array += DC395x_MAX_SG_LISTENTRY;
-	}
-
 	return 0;
 }
 

drivers/video/fbmem.c

@@ -628,7 +628,7 @@ fb_pan_display(struct fb_info *info, struct fb_var_screeninfo *var)
 int
 fb_set_var(struct fb_info *info, struct fb_var_screeninfo *var)
 {
-	int err;
+	int err, flags = info->flags;
 
 	if (var->activate & FB_ACTIVATE_INV_MODE) {
 		struct fb_videomode mode1, mode2;
@@ -682,7 +682,7 @@ fb_set_var(struct fb_info *info, struct fb_var_screeninfo *var)
 			    !list_empty(&info->modelist))
 				err = fb_add_videomode(&mode, &info->modelist);
 
-			if (!err && info->flags & FBINFO_MISC_USEREVENT) {
+			if (!err && (flags & FBINFO_MISC_USEREVENT)) {
 				struct fb_event event;
 
 				info->flags &= ~FBINFO_MISC_USEREVENT;

drivers/video/intelfb/intelfbdrv.c

@@ -583,23 +583,6 @@ intelfb_pci_register(struct pci_dev *pdev, const struct pci_device_id *ent)
 		return -ENODEV;
 	}
 
-	/* Map the fb and MMIO regions */
-	dinfo->aperture.virtual = (u8 __iomem *)ioremap_nocache
-		(dinfo->aperture.physical, dinfo->aperture.size);
-	if (!dinfo->aperture.virtual) {
-		ERR_MSG("Cannot remap FB region.\n");
-		cleanup(dinfo);
-		return -ENODEV;
-	}
-	dinfo->mmio_base =
-		(u8 __iomem *)ioremap_nocache(dinfo->mmio_base_phys,
-					       INTEL_REG_SIZE);
-	if (!dinfo->mmio_base) {
-		ERR_MSG("Cannot remap MMIO region.\n");
-		cleanup(dinfo);
-		return -ENODEV;
-	}
-
 	/* Get the chipset info. */
 	dinfo->pci_chipset = pdev->device;
 
@@ -630,9 +613,15 @@ intelfb_pci_register(struct pci_dev *pdev, const struct pci_device_id *ent)
 		dinfo->accel = 0;
 	}
 
+	if (MB(voffset) < stolen_size)
+		offset = (stolen_size >> 12);
+	else
+		offset = ROUND_UP_TO_PAGE(MB(voffset))/GTT_PAGE_SIZE;
+
 	/* Framebuffer parameters - Use all the stolen memory if >= vram */
-	if (ROUND_UP_TO_PAGE(stolen_size) >= MB(vram)) {
+	if (ROUND_UP_TO_PAGE(stolen_size) >= ((offset << 12) +  MB(vram))) {
 		dinfo->fb.size = ROUND_UP_TO_PAGE(stolen_size);
+		dinfo->fb.offset = 0;
 		dinfo->fbmem_gart = 0;
 	} else {
 		dinfo->fb.size =  MB(vram);
@@ -663,11 +652,6 @@ intelfb_pci_register(struct pci_dev *pdev, const struct pci_device_id *ent)
 		return -ENODEV;
 	}
 
-	if (MB(voffset) < stolen_size)
-		offset = (stolen_size >> 12);
-	else
-		offset = ROUND_UP_TO_PAGE(MB(voffset))/GTT_PAGE_SIZE;
-
 	/* set the mem offsets - set them after the already used pages */
 	if (dinfo->accel) {
 		dinfo->ring.offset = offset + gtt_info.current_memory;
@@ -682,6 +666,26 @@ intelfb_pci_register(struct pci_dev *pdev, const struct pci_device_id *ent)
 			+ (dinfo->cursor.size >> 12);
 	}
 
+	/* Map the fb and MMIO regions */
+	/* ioremap only up to the end of used aperture */
+	dinfo->aperture.virtual = (u8 __iomem *)ioremap_nocache
+		(dinfo->aperture.physical, (dinfo->fb.offset << 12)
+		 + dinfo->fb.size);
+	if (!dinfo->aperture.virtual) {
+		ERR_MSG("Cannot remap FB region.\n");
+		cleanup(dinfo);
+		return -ENODEV;
+	}
+
+	dinfo->mmio_base =
+		(u8 __iomem *)ioremap_nocache(dinfo->mmio_base_phys,
+					       INTEL_REG_SIZE);
+	if (!dinfo->mmio_base) {
+		ERR_MSG("Cannot remap MMIO region.\n");
+		cleanup(dinfo);
+		return -ENODEV;
+	}
+
 	/* Allocate memories (which aren't stolen) */
 	if (dinfo->accel) {
 		if (!(dinfo->gtt_ring_mem =

drivers/w1/w1.c

@@ -593,7 +593,7 @@ void w1_search(struct w1_master *dev, w1_slave_found_callback cb)
 		 * Return 0 - device(s) present, 1 - no devices present.
 		 */
 		if (w1_reset_bus(dev)) {
-			dev_info(&dev->dev, "No devices present on the wire.\n");
+			dev_dbg(&dev->dev, "No devices present on the wire.\n");
 			break;
 		}
 

fs/cifs/CHANGES

+Version 1.35
+------------
+Add writepage performance improvements.  Fix path name conversions
+for long filenames on mounts which were done with "mapchars" mount option
+specified.
+
 Version 1.34
 ------------
 Fix error mapping of the TOO_MANY_LINKS (hardlinks) case.

fs/cifs/cifssmb.c

@@ -2602,6 +2602,9 @@ int CIFSFindNext(const int xid, struct cifsTconInfo *tcon,
 	if(name_len < PATH_MAX) {
 		memcpy(pSMB->ResumeFileName, psrch_inf->presume_name, name_len);
 		byte_count += name_len;
+		/* 14 byte parm len above enough for 2 byte null terminator */
+		pSMB->ResumeFileName[name_len] = 0;
+		pSMB->ResumeFileName[name_len+1] = 0;
 	} else {
 		rc = -EINVAL;
 		goto FNext2_err_exit;

fs/cifs/misc.c

@@ -611,6 +611,7 @@ cifsConvertToUCS(__le16 * target, const char *source, int maxlen,
 		src_char = source[i];
 		switch (src_char) {
 			case 0:
+				target[j] = 0;
 				goto ctoUCS_out;
 			case ':':
 				target[j] = cpu_to_le16(UNI_COLON);

fs/inotify.c

@@ -402,7 +402,7 @@ static struct inotify_watch *create_watch(struct inotify_device *dev,
 		return ERR_PTR(ret);
 	}
 
-	dev->last_wd = ret;
+	dev->last_wd = watch->wd;
 	watch->mask = mask;
 	atomic_set(&watch->count, 0);
 	INIT_LIST_HEAD(&watch->d_list);

fs/namei.c

@@ -2216,7 +2216,8 @@ int vfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 		error = vfs_rename_other(old_dir,old_dentry,new_dir,new_dentry);
 	if (!error) {
 		const char *new_name = old_dentry->d_name.name;
-		fsnotify_move(old_dir, new_dir, old_name, new_name, is_dir, new_dentry->d_inode);
+		fsnotify_move(old_dir, new_dir, old_name, new_name, is_dir,
+			      new_dentry->d_inode, old_dentry->d_inode);
 	}
 	fsnotify_oldname_free(old_name);
 

fs/reiserfs/namei.c

@@ -593,6 +593,9 @@ static int new_inode_init(struct inode *inode, struct inode *dir, int mode)
 	 */
 	inode->i_uid = current->fsuid;
 	inode->i_mode = mode;
+	/* Make inode invalid - just in case we are going to drop it before
+	 * the initialization happens */
+	INODE_PKEY(inode)->k_objectid = 0;
 
 	if (dir->i_mode & S_ISGID) {
 		inode->i_gid = dir->i_gid;

include/asm-i386/pci.h

@@ -18,9 +18,11 @@ extern unsigned int pcibios_assign_all_busses(void);
 #define pcibios_scan_all_fns(a, b)	0
 
 extern unsigned long pci_mem_start;
-#define PCIBIOS_MIN_IO		0x4000
+#define PCIBIOS_MIN_IO		0x1000
 #define PCIBIOS_MIN_MEM		(pci_mem_start)
 
+#define PCIBIOS_MIN_CARDBUS_IO	0x4000
+
 void pcibios_config_init(void);
 struct pci_bus * pcibios_scan_root(int bus);
 

include/asm-sh/unistd.h

@@ -406,7 +406,7 @@ register long __sc6 __asm__ ("r6") = (long) arg3; \
 register long __sc7 __asm__ ("r7") = (long) arg4; \
 register long __sc0 __asm__ ("r0") = (long) arg5; \
 register long __sc1 __asm__ ("r1") = (long) arg6; \
-__asm__ __volatile__ ("trapa	#0x15" \
+__asm__ __volatile__ ("trapa	#0x16" \
 	: "=z" (__sc0) \
 	: "0" (__sc0), "r" (__sc4), "r" (__sc5), "r" (__sc6), "r" (__sc7),  \
 	  "r" (__sc3), "r" (__sc1) \

include/asm-x86_64/pci.h

@@ -22,9 +22,11 @@ extern unsigned int pcibios_assign_all_busses(void);
 extern int no_iommu, force_iommu;
 
 extern unsigned long pci_mem_start;
-#define PCIBIOS_MIN_IO		0x4000
+#define PCIBIOS_MIN_IO		0x1000
 #define PCIBIOS_MIN_MEM		(pci_mem_start)
 
+#define PCIBIOS_MIN_CARDBUS_IO	0x4000
+
 void pcibios_config_init(void);
 struct pci_bus * pcibios_scan_root(int bus);
 extern int (*pci_config_read)(int seg, int bus, int dev, int fn, int reg, int len, u32 *value);

include/linux/fsnotify.h

@@ -21,7 +21,7 @@
  */
 static inline void fsnotify_move(struct inode *old_dir, struct inode *new_dir,
 				 const char *old_name, const char *new_name,
-				 int isdir, struct inode *target)
+				 int isdir, struct inode *target, struct inode *source)
 {
 	u32 cookie = inotify_get_cookie();
 
@@ -41,6 +41,10 @@ static inline void fsnotify_move(struct inode *old_dir, struct inode *new_dir,
 		inotify_inode_queue_event(target, IN_DELETE_SELF, 0, NULL);
 		inotify_inode_is_dead(target);
 	}
+
+	if (source) {
+		inotify_inode_queue_event(source, IN_MOVE_SELF, 0, NULL);
+	}
 }
 
 /*

include/linux/inotify.h

@@ -35,6 +35,7 @@ struct inotify_event {
 #define IN_CREATE		0x00000100	/* Subfile was created */
 #define IN_DELETE		0x00000200	/* Subfile was deleted */
 #define IN_DELETE_SELF		0x00000400	/* Self was deleted */
+#define IN_MOVE_SELF		0x00000800	/* Self was moved */
 
 /* the following are legal events.  they are sent as needed to any watch */
 #define IN_UNMOUNT		0x00002000	/* Backing fs was unmounted */
@@ -56,7 +57,8 @@ struct inotify_event {
  */
 #define IN_ALL_EVENTS	(IN_ACCESS | IN_MODIFY | IN_ATTRIB | IN_CLOSE_WRITE | \
 			 IN_CLOSE_NOWRITE | IN_OPEN | IN_MOVED_FROM | \
-			 IN_MOVED_TO | IN_DELETE | IN_CREATE | IN_DELETE_SELF)
+			 IN_MOVED_TO | IN_DELETE | IN_CREATE | IN_DELETE_SELF | \
+			 IN_MOVE_SELF)
 
 #ifdef __KERNEL__
 

include/linux/netpoll.h

@@ -9,6 +9,7 @@
 
 #include <linux/netdevice.h>
 #include <linux/interrupt.h>
+#include <linux/rcupdate.h>
 #include <linux/list.h>
 
 struct netpoll;
@@ -26,6 +27,7 @@ struct netpoll {
 struct netpoll_info {
 	spinlock_t poll_lock;
 	int poll_owner;
+	int tries;
 	int rx_flags;
 	spinlock_t rx_lock;
 	struct netpoll *rx_np; /* netpoll that registered an rx_hook */
@@ -60,25 +62,31 @@ static inline int netpoll_rx(struct sk_buff *skb)
 	return ret;
 }
 
-static inline void netpoll_poll_lock(struct net_device *dev)
+static inline void *netpoll_poll_lock(struct net_device *dev)
 {
+	rcu_read_lock(); /* deal with race on ->npinfo */
 	if (dev->npinfo) {
 		spin_lock(&dev->npinfo->poll_lock);
 		dev->npinfo->poll_owner = smp_processor_id();
+		return dev->npinfo;
 	}
+	return NULL;
 }
 
-static inline void netpoll_poll_unlock(struct net_device *dev)
+static inline void netpoll_poll_unlock(void *have)
 {
-	if (dev->npinfo) {
-		dev->npinfo->poll_owner = -1;
-		spin_unlock(&dev->npinfo->poll_lock);
+	struct netpoll_info *npi = have;
+
+	if (npi) {
+		npi->poll_owner = -1;
+		spin_unlock(&npi->poll_lock);
 	}
+	rcu_read_unlock();
 }
 
 #else
 #define netpoll_rx(a) 0
-#define netpoll_poll_lock(a)
+#define netpoll_poll_lock(a) 0
 #define netpoll_poll_unlock(a)
 #endif
 

include/linux/skbuff.h

@@ -255,7 +255,7 @@ struct sk_buff {
 				nohdr:1;
 				/* 3 bits spare */
 	__u8			pkt_type;
-	__u16			protocol;
+	__be16			protocol;
 
 	void			(*destructor)(struct sk_buff *skb);
 #ifdef CONFIG_NETFILTER

net/core/dev.c

@@ -1696,6 +1696,7 @@ static void net_rx_action(struct softirq_action *h)
 	struct softnet_data *queue = &__get_cpu_var(softnet_data);
 	unsigned long start_time = jiffies;
 	int budget = netdev_budget;
+	void *have;
 
 	local_irq_disable();
 
@@ -1709,10 +1710,10 @@ static void net_rx_action(struct softirq_action *h)
 
 		dev = list_entry(queue->poll_list.next,
 				 struct net_device, poll_list);
-		netpoll_poll_lock(dev);
+		have = netpoll_poll_lock(dev);
 
 		if (dev->quota <= 0 || dev->poll(dev, &budget)) {
-			netpoll_poll_unlock(dev);
+			netpoll_poll_unlock(have);
 			local_irq_disable();
 			list_del(&dev->poll_list);
 			list_add_tail(&dev->poll_list, &queue->poll_list);
@@ -1721,7 +1722,7 @@ static void net_rx_action(struct softirq_action *h)
 			else
 				dev->quota = dev->weight;
 		} else {
-			netpoll_poll_unlock(dev);
+			netpoll_poll_unlock(have);
 			dev_put(dev);
 			local_irq_disable();
 		}

net/core/netpoll.c

@@ -33,6 +33,7 @@
 #define MAX_UDP_CHUNK 1460
 #define MAX_SKBS 32
 #define MAX_QUEUE_DEPTH (MAX_SKBS / 2)
+#define MAX_RETRIES 20000
 
 static DEFINE_SPINLOCK(skb_list_lock);
 static int nr_skbs;
@@ -248,14 +249,14 @@ static void netpoll_send_skb(struct netpoll *np, struct sk_buff *skb)
 	int status;
 	struct netpoll_info *npinfo;
 
-repeat:
-	if(!np || !np->dev || !netif_running(np->dev)) {
+	if (!np || !np->dev || !netif_running(np->dev)) {
 		__kfree_skb(skb);
 		return;
 	}
 
-	/* avoid recursion */
 	npinfo = np->dev->npinfo;
+
+	/* avoid recursion */
 	if (npinfo->poll_owner == smp_processor_id() ||
 	    np->dev->xmit_lock_owner == smp_processor_id()) {
 		if (np->drop)
@@ -265,6 +266,8 @@ repeat:
 		return;
 	}
 
+	do {
+		npinfo->tries--;
 		spin_lock(&np->dev->xmit_lock);
 		np->dev->xmit_lock_owner = smp_processor_id();
 
@@ -275,20 +278,25 @@ repeat:
 		if (netif_queue_stopped(np->dev)) {
 			np->dev->xmit_lock_owner = -1;
 			spin_unlock(&np->dev->xmit_lock);
-
 			netpoll_poll(np);
-		goto repeat;
+			udelay(50);
+			continue;
 		}
 
 		status = np->dev->hard_start_xmit(skb, np->dev);
 		np->dev->xmit_lock_owner = -1;
 		spin_unlock(&np->dev->xmit_lock);
 
+		/* success */
+		if(!status) {
+			npinfo->tries = MAX_RETRIES; /* reset */
+			return;
+		}
+
 		/* transmit busy */
-	if(status) {
 		netpoll_poll(np);
-		goto repeat;
-	}
+		udelay(50);
+	} while (npinfo->tries > 0);
 }
 
 void netpoll_send_udp(struct netpoll *np, const char *msg, int len)
@@ -349,15 +357,11 @@ static void arp_reply(struct sk_buff *skb)
 	unsigned char *arp_ptr;
 	int size, type = ARPOP_REPLY, ptype = ETH_P_ARP;
 	u32 sip, tip;
-	unsigned long flags;
 	struct sk_buff *send_skb;
 	struct netpoll *np = NULL;
 
-	spin_lock_irqsave(&npinfo->rx_lock, flags);
 	if (npinfo->rx_np && npinfo->rx_np->dev == skb->dev)
 		np = npinfo->rx_np;
-	spin_unlock_irqrestore(&npinfo->rx_lock, flags);
-
 	if (!np)
 		return;
 
@@ -639,9 +643,11 @@ int netpoll_setup(struct netpoll *np)
 		if (!npinfo)
 			goto release;
 
+		npinfo->rx_flags = 0;
 		npinfo->rx_np = NULL;
 		npinfo->poll_lock = SPIN_LOCK_UNLOCKED;
 		npinfo->poll_owner = -1;
+		npinfo->tries = MAX_RETRIES;
 		npinfo->rx_lock = SPIN_LOCK_UNLOCKED;
 	} else
 		npinfo = ndev->npinfo;
@@ -718,9 +724,16 @@ int netpoll_setup(struct netpoll *np)
 		npinfo->rx_np = np;
 		spin_unlock_irqrestore(&npinfo->rx_lock, flags);
 	}
+
+	/* fill up the skb queue */
+	refill_skbs();
+
 	/* last thing to do is link it to the net device structure */
 	ndev->npinfo = npinfo;
 
+	/* avoid racing with NAPI reading npinfo */
+	synchronize_rcu();
+
 	return 0;
 
  release: