From a03fa955576af50df80bec9127b46ef57e0877c0 Mon Sep 17 00:00:00 2001
From: "rajesh.shah@intel.com" <rajesh.shah@intel.com>
Date: Thu, 2 Jun 2005 15:41:48 -0700
Subject: [PATCH] PCI: Increase the number of PCI bus resources

This patch increases the number of resource pointers in the
pci_bus structure. This is needed to store >4 resource ranges
for host bridges and transparent PCI bridges. With this change,
all PCI buses will have more resource pointers, but most PCI
buses will only use the first 3 or 4, the remaining being NULL.
The PCI core already deals with this correctly.

Signed-off-by: Rajesh Shah <rajesh.shah@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 include/linux/pci.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/pci.h b/include/linux/pci.h
index 66798b46f308..a46cabfd08c8 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -586,7 +586,7 @@ struct pci_dev {
 #define PCI_NUM_RESOURCES 11
 
 #ifndef PCI_BUS_NUM_RESOURCES
-#define PCI_BUS_NUM_RESOURCES 4
+#define PCI_BUS_NUM_RESOURCES 8
 #endif
   
 #define PCI_REGION_FLAG_MASK 0x0fU	/* These bits of resource flags tell us the PCI region flags */
-- 
cgit v1.2.3


From 75865858971add95809c5c9cd35dc4cfba08e33b Mon Sep 17 00:00:00 2001
From: Greg Kroah-Hartman <gregkh@suse.de>
Date: Thu, 30 Jun 2005 02:18:12 -0700
Subject: [PATCH] PCI: clean up dynamic pci id logic

The dynamic pci id logic has been bothering me for a while, and now that
I started to look into how to move some of this to the driver core, I
thought it was time to clean it all up.

It ends up making the code smaller, and easier to follow, and fixes a
few bugs at the same time (dynamic ids were not being matched
everywhere, and so could be missed on some call paths for new devices,
semaphore not needed to be grabbed when adding a new id and calling the
driver core, etc.)

I also renamed the function pci_match_device() to pci_match_id() as
that's what it really does.

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---
 arch/i386/kernel/cpu/cpufreq/gx-suspmod.c |   2 +-
 drivers/char/hw_random.c                  |   2 +-
 drivers/char/watchdog/i8xx_tco.c          |   2 +-
 drivers/ide/setup-pci.c                   |   2 +-
 drivers/parport/parport_pc.c              |   2 +-
 drivers/pci/pci-driver.c                  | 196 +++++++++++-------------------
 include/linux/pci-dynids.h                |  18 ---
 include/linux/pci.h                       |   3 +-
 sound/pci/bt87x.c                         |   2 +-
 9 files changed, 79 insertions(+), 150 deletions(-)
 delete mode 100644 include/linux/pci-dynids.h

(limited to 'include/linux')

diff --git a/arch/i386/kernel/cpu/cpufreq/gx-suspmod.c b/arch/i386/kernel/cpu/cpufreq/gx-suspmod.c
index 1a49adb1f4a6..e86ea486c311 100644
--- a/arch/i386/kernel/cpu/cpufreq/gx-suspmod.c
+++ b/arch/i386/kernel/cpu/cpufreq/gx-suspmod.c
@@ -190,7 +190,7 @@ static __init struct pci_dev *gx_detect_chipset(void)
 
 	/* detect which companion chip is used */
 	while ((gx_pci = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, gx_pci)) != NULL) {
-		if ((pci_match_device (gx_chipset_tbl, gx_pci)) != NULL) {
+		if ((pci_match_id(gx_chipset_tbl, gx_pci)) != NULL) {
 			return gx_pci;
 		}
 	}
diff --git a/drivers/char/hw_random.c b/drivers/char/hw_random.c
index 7e6ac14c2450..3480535a09c5 100644
--- a/drivers/char/hw_random.c
+++ b/drivers/char/hw_random.c
@@ -579,7 +579,7 @@ static int __init rng_init (void)
 
 	/* Probe for Intel, AMD RNGs */
 	for_each_pci_dev(pdev) {
-		ent = pci_match_device (rng_pci_tbl, pdev);
+		ent = pci_match_id(rng_pci_tbl, pdev);
 		if (ent) {
 			rng_ops = &rng_vendor_ops[ent->driver_data];
 			goto match;
diff --git a/drivers/char/watchdog/i8xx_tco.c b/drivers/char/watchdog/i8xx_tco.c
index b14d642439ed..5d07ee59679d 100644
--- a/drivers/char/watchdog/i8xx_tco.c
+++ b/drivers/char/watchdog/i8xx_tco.c
@@ -401,7 +401,7 @@ static unsigned char __init i8xx_tco_getdevice (void)
 	 */
 
 	while ((dev = pci_find_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) {
-		if (pci_match_device(i8xx_tco_pci_tbl, dev)) {
+		if (pci_match_id(i8xx_tco_pci_tbl, dev)) {
 			i8xx_tco_pci = dev;
 			break;
 		}
diff --git a/drivers/ide/setup-pci.c b/drivers/ide/setup-pci.c
index e501675ad72e..77da827b2898 100644
--- a/drivers/ide/setup-pci.c
+++ b/drivers/ide/setup-pci.c
@@ -847,7 +847,7 @@ static int __init ide_scan_pcidev(struct pci_dev *dev)
 		d = list_entry(l, struct pci_driver, node);
 		if(d->id_table)
 		{
-			const struct pci_device_id *id = pci_match_device(d->id_table, dev);
+			const struct pci_device_id *id = pci_match_id(d->id_table, dev);
 			if(id != NULL)
 			{
 				if(d->probe(dev, id) >= 0)
diff --git a/drivers/parport/parport_pc.c b/drivers/parport/parport_pc.c
index 80edfa3abd29..4598c6a9212d 100644
--- a/drivers/parport/parport_pc.c
+++ b/drivers/parport/parport_pc.c
@@ -3008,7 +3008,7 @@ static int __init parport_pc_init_superio (int autoirq, int autodma)
 	int ret = 0;
 
 	while ((pdev = pci_find_device(PCI_ANY_ID, PCI_ANY_ID, pdev)) != NULL) {
-		id = pci_match_device (parport_pc_pci_tbl, pdev);
+		id = pci_match_id(parport_pc_pci_tbl, pdev);
 		if (id == NULL || id->driver_data >= last_sio)
 			continue;
 
diff --git a/drivers/pci/pci-driver.c b/drivers/pci/pci-driver.c
index e65bf2b395aa..aac6de9568e5 100644
--- a/drivers/pci/pci-driver.c
+++ b/drivers/pci/pci-driver.c
@@ -7,7 +7,6 @@
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/device.h>
-#include <linux/pci-dynids.h>
 #include "pci.h"
 
 /*
@@ -19,35 +18,11 @@
  */
 
 #ifdef CONFIG_HOTPLUG
-/**
- * pci_device_probe_dynamic()
- *
- * Walk the dynamic ID list looking for a match.
- * returns 0 and sets pci_dev->driver when drv claims pci_dev, else error.
- */
-static int
-pci_device_probe_dynamic(struct pci_driver *drv, struct pci_dev *pci_dev)
-{
-	int error = -ENODEV;
-	struct list_head *pos;
-	struct dynid *dynid;
 
-	spin_lock(&drv->dynids.lock);
-	list_for_each(pos, &drv->dynids.list) {
-		dynid = list_entry(pos, struct dynid, node);
-		if (pci_match_one_device(&dynid->id, pci_dev)) {
-			spin_unlock(&drv->dynids.lock);
-			error = drv->probe(pci_dev, &dynid->id);
-			if (error >= 0) {
-				pci_dev->driver = drv;
-				return 0;
-			}
-			return error;
-		}
-	}
-	spin_unlock(&drv->dynids.lock);
-	return error;
-}
+struct pci_dynid {
+	struct list_head node;
+	struct pci_device_id id;
+};
 
 /**
  * store_new_id
@@ -58,8 +33,7 @@ pci_device_probe_dynamic(struct pci_driver *drv, struct pci_dev *pci_dev)
 static inline ssize_t
 store_new_id(struct device_driver *driver, const char *buf, size_t count)
 {
-	struct dynid *dynid;
-	struct bus_type * bus;
+	struct pci_dynid *dynid;
 	struct pci_driver *pdrv = to_pci_driver(driver);
 	__u32 vendor=PCI_ANY_ID, device=PCI_ANY_ID, subvendor=PCI_ANY_ID,
 		subdevice=PCI_ANY_ID, class=0, class_mask=0;
@@ -91,37 +65,22 @@ store_new_id(struct device_driver *driver, const char *buf, size_t count)
 	list_add_tail(&pdrv->dynids.list, &dynid->node);
 	spin_unlock(&pdrv->dynids.lock);
 
-	bus = get_bus(pdrv->driver.bus);
-	if (bus) {
-		if (get_driver(&pdrv->driver)) {
-			down_write(&bus->subsys.rwsem);
-			driver_attach(&pdrv->driver);
-			up_write(&bus->subsys.rwsem);
-			put_driver(&pdrv->driver);
-		}
-		put_bus(bus);
+	if (get_driver(&pdrv->driver)) {
+		driver_attach(&pdrv->driver);
+		put_driver(&pdrv->driver);
 	}
 
 	return count;
 }
-
 static DRIVER_ATTR(new_id, S_IWUSR, NULL, store_new_id);
-static inline void
-pci_init_dynids(struct pci_dynids *dynids)
-{
-	spin_lock_init(&dynids->lock);
-	INIT_LIST_HEAD(&dynids->list);
-}
 
 static void
 pci_free_dynids(struct pci_driver *drv)
 {
-	struct list_head *pos, *n;
-	struct dynid *dynid;
+	struct pci_dynid *dynid, *n;
 
 	spin_lock(&drv->dynids.lock);
-	list_for_each_safe(pos, n, &drv->dynids.list) {
-		dynid = list_entry(pos, struct dynid, node);
+	list_for_each_entry_safe(dynid, n, &drv->dynids.list, node) {
 		list_del(&dynid->node);
 		kfree(dynid);
 	}
@@ -138,83 +97,70 @@ pci_create_newid_file(struct pci_driver *drv)
 	return error;
 }
 
-static int
-pci_bus_match_dynids(const struct pci_dev *pci_dev, struct pci_driver *pci_drv)
-{
-	struct list_head *pos;
-	struct dynid *dynid;
-
-	spin_lock(&pci_drv->dynids.lock);
-	list_for_each(pos, &pci_drv->dynids.list) {
-		dynid = list_entry(pos, struct dynid, node);
-		if (pci_match_one_device(&dynid->id, pci_dev)) {
-			spin_unlock(&pci_drv->dynids.lock);
-			return 1;
-		}
-	}
-	spin_unlock(&pci_drv->dynids.lock);
-	return 0;
-}
-
 #else /* !CONFIG_HOTPLUG */
-static inline int pci_device_probe_dynamic(struct pci_driver *drv, struct pci_dev *pci_dev)
-{
-	return -ENODEV;
-}
-static inline void pci_init_dynids(struct pci_dynids *dynids) {}
 static inline void pci_free_dynids(struct pci_driver *drv) {}
 static inline int pci_create_newid_file(struct pci_driver *drv)
 {
 	return 0;
 }
-static inline int pci_bus_match_dynids(const struct pci_dev *pci_dev, struct pci_driver *pci_drv)
-{
-	return 0;
-}
 #endif
 
 /**
- * pci_match_device - Tell if a PCI device structure has a matching
- *                    PCI device id structure
+ * pci_match_id - See if a pci device matches a given pci_id table
  * @ids: array of PCI device id structures to search in
- * @dev: the PCI device structure to match against
- * 
+ * @dev: the PCI device structure to match against.
+ *
  * Used by a driver to check whether a PCI device present in the
- * system is in its list of supported devices.Returns the matching
+ * system is in its list of supported devices.  Returns the matching
  * pci_device_id structure or %NULL if there is no match.
+ *
+ * Depreciated, don't use this as it will not catch any dynamic ids
+ * that a driver might want to check for.
  */
-const struct pci_device_id *
-pci_match_device(const struct pci_device_id *ids, const struct pci_dev *dev)
+const struct pci_device_id *pci_match_id(const struct pci_device_id *ids,
+					 struct pci_dev *dev)
 {
-	while (ids->vendor || ids->subvendor || ids->class_mask) {
-		if (pci_match_one_device(ids, dev))
-			return ids;
-		ids++;
+	if (ids) {
+		while (ids->vendor || ids->subvendor || ids->class_mask) {
+			if (pci_match_one_device(ids, dev))
+				return ids;
+			ids++;
+		}
 	}
 	return NULL;
 }
 
 /**
- * pci_device_probe_static()
- * 
- * returns 0 and sets pci_dev->driver when drv claims pci_dev, else error.
+ * pci_match_device - Tell if a PCI device structure has a matching
+ *                    PCI device id structure
+ * @ids: array of PCI device id structures to search in
+ * @dev: the PCI device structure to match against
+ * @drv: the PCI driver to match against
+ *
+ * Used by a driver to check whether a PCI device present in the
+ * system is in its list of supported devices.  Returns the matching
+ * pci_device_id structure or %NULL if there is no match.
  */
-static int
-pci_device_probe_static(struct pci_driver *drv, struct pci_dev *pci_dev)
-{		   
-	int error = -ENODEV;
+const struct pci_device_id *pci_match_device(struct pci_driver *drv,
+					     struct pci_dev *dev)
+{
 	const struct pci_device_id *id;
+	struct pci_dynid *dynid;
 
-	if (!drv->id_table)
-		return error;
-	id = pci_match_device(drv->id_table, pci_dev);
+	id = pci_match_id(drv->id_table, dev);
 	if (id)
-		error = drv->probe(pci_dev, id);
-	if (error >= 0) {
-		pci_dev->driver = drv;
-		error = 0;
+		return id;
+
+	/* static ids didn't match, lets look at the dynamic ones */
+	spin_lock(&drv->dynids.lock);
+	list_for_each_entry(dynid, &drv->dynids.list, node) {
+		if (pci_match_one_device(&dynid->id, dev)) {
+			spin_unlock(&drv->dynids.lock);
+			return &dynid->id;
+		}
 	}
-	return error;
+	spin_unlock(&drv->dynids.lock);
+	return NULL;
 }
 
 /**
@@ -225,13 +171,20 @@ pci_device_probe_static(struct pci_driver *drv, struct pci_dev *pci_dev)
  */
 static int
 __pci_device_probe(struct pci_driver *drv, struct pci_dev *pci_dev)
-{		   
+{
+	const struct pci_device_id *id;
 	int error = 0;
 
 	if (!pci_dev->driver && drv->probe) {
-		error = pci_device_probe_static(drv, pci_dev);
-		if (error == -ENODEV)
-			error = pci_device_probe_dynamic(drv, pci_dev);
+		error = -ENODEV;
+
+		id = pci_match_device(drv, pci_dev);
+		if (id)
+			error = drv->probe(pci_dev, id);
+		if (error >= 0) {
+			pci_dev->driver = drv;
+			error = 0;
+		}
 	}
 	return error;
 }
@@ -371,12 +324,6 @@ static struct kobj_type pci_driver_kobj_type = {
 	.sysfs_ops = &pci_driver_sysfs_ops,
 };
 
-static int
-pci_populate_driver_dir(struct pci_driver *drv)
-{
-	return pci_create_newid_file(drv);
-}
-
 /**
  * pci_register_driver - register a new pci driver
  * @drv: the driver structure to register
@@ -401,13 +348,15 @@ int pci_register_driver(struct pci_driver *drv)
 		drv->driver.shutdown = pci_device_shutdown;
 	drv->driver.owner = drv->owner;
 	drv->driver.kobj.ktype = &pci_driver_kobj_type;
-	pci_init_dynids(&drv->dynids);
+
+	spin_lock_init(&drv->dynids.lock);
+	INIT_LIST_HEAD(&drv->dynids.list);
 
 	/* register with core */
 	error = driver_register(&drv->driver);
 
 	if (!error)
-		pci_populate_driver_dir(drv);
+		error = pci_create_newid_file(drv);
 
 	return error;
 }
@@ -463,21 +412,17 @@ pci_dev_driver(const struct pci_dev *dev)
  * system is in its list of supported devices.Returns the matching
  * pci_device_id structure or %NULL if there is no match.
  */
-static int pci_bus_match(struct device * dev, struct device_driver * drv) 
+static int pci_bus_match(struct device *dev, struct device_driver *drv)
 {
-	const struct pci_dev * pci_dev = to_pci_dev(dev);
-	struct pci_driver * pci_drv = to_pci_driver(drv);
-	const struct pci_device_id * ids = pci_drv->id_table;
+	struct pci_dev *pci_dev = to_pci_dev(dev);
+	struct pci_driver *pci_drv = to_pci_driver(drv);
 	const struct pci_device_id *found_id;
 
-	if (!ids)
-		return 0;
-
-	found_id = pci_match_device(ids, pci_dev);
+	found_id = pci_match_device(pci_drv, pci_dev);
 	if (found_id)
 		return 1;
 
-	return pci_bus_match_dynids(pci_dev, pci_drv);
+	return 0;
 }
 
 /**
@@ -536,6 +481,7 @@ static int __init pci_driver_init(void)
 
 postcore_initcall(pci_driver_init);
 
+EXPORT_SYMBOL(pci_match_id);
 EXPORT_SYMBOL(pci_match_device);
 EXPORT_SYMBOL(pci_register_driver);
 EXPORT_SYMBOL(pci_unregister_driver);
diff --git a/include/linux/pci-dynids.h b/include/linux/pci-dynids.h
deleted file mode 100644
index 183b6b0de81c..000000000000
--- a/include/linux/pci-dynids.h
+++ /dev/null
@@ -1,18 +0,0 @@
-/*
- *	PCI defines and function prototypes
- *	Copyright 2003 Dell Inc.
- *        by Matt Domsch <Matt_Domsch@dell.com>
- */
-
-#ifndef LINUX_PCI_DYNIDS_H
-#define LINUX_PCI_DYNIDS_H
-
-#include <linux/list.h>
-#include <linux/mod_devicetable.h>
-
-struct dynid {
-	struct list_head        node;
-	struct pci_device_id    id;
-};
-
-#endif
diff --git a/include/linux/pci.h b/include/linux/pci.h
index a46cabfd08c8..7ac14961ba22 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -860,7 +860,8 @@ int pci_register_driver(struct pci_driver *);
 void pci_unregister_driver(struct pci_driver *);
 void pci_remove_behind_bridge(struct pci_dev *);
 struct pci_driver *pci_dev_driver(const struct pci_dev *);
-const struct pci_device_id *pci_match_device(const struct pci_device_id *ids, const struct pci_dev *dev);
+const struct pci_device_id *pci_match_device(struct pci_driver *drv, struct pci_dev *dev);
+const struct pci_device_id *pci_match_id(const struct pci_device_id *ids, struct pci_dev *dev);
 int pci_scan_bridge(struct pci_bus *bus, struct pci_dev * dev, int max, int pass);
 
 /* kmem_cache style wrapper around pci_alloc_consistent() */
diff --git a/sound/pci/bt87x.c b/sound/pci/bt87x.c
index defdc5a459f0..909fef8903cb 100644
--- a/sound/pci/bt87x.c
+++ b/sound/pci/bt87x.c
@@ -804,7 +804,7 @@ static int __devinit snd_bt87x_detect_card(struct pci_dev *pci)
 	int i;
 	const struct pci_device_id *supported;
 
-	supported = pci_match_device(snd_bt87x_ids, pci);
+	supported = pci_match_device(driver, pci);
 	if (supported)
 		return supported->driver_data;
 
-- 
cgit v1.2.3


From 21e2c01dc3e38d466eda5871645878d2c3a33261 Mon Sep 17 00:00:00 2001
From: Rob Punkunus <rpunkunus@nvidia.com>
Date: Sun, 3 Jul 2005 17:37:18 +0200
Subject: [PATCH] amd74xx: support MCP55 device IDs

From: Rob Punkunus <rpunkunus@nvidia.com>

Rob Punkunus recently submitted a patch to enable support for MCP51/MCP55 in
the amd74xx driver. This patch was whitespace-corrupted and didn't apply to
2.6.12 since MCP51 support was merged in the 2.6.12-rc series.

Gentoo would like to support this hardware for our upcoming release media, so
I fixed the patch, and here it is :)

Signed-off-by: Daniel Drake <dsd@gentoo.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@elka.pw.edu.pl>
---
 drivers/ide/pci/amd74xx.c | 3 +++
 include/linux/pci_ids.h   | 1 +
 2 files changed, 4 insertions(+)

(limited to 'include/linux')

diff --git a/drivers/ide/pci/amd74xx.c b/drivers/ide/pci/amd74xx.c
index 65eab9b63a79..844a6c9fb949 100644
--- a/drivers/ide/pci/amd74xx.c
+++ b/drivers/ide/pci/amd74xx.c
@@ -73,6 +73,7 @@ static struct amd_ide_chip {
 	{ PCI_DEVICE_ID_NVIDIA_NFORCE_CK804_IDE,	0x50, AMD_UDMA_133 },
 	{ PCI_DEVICE_ID_NVIDIA_NFORCE_MCP04_IDE,	0x50, AMD_UDMA_133 },
 	{ PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_IDE,	0x50, AMD_UDMA_133 },
+	{ PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_IDE,	0x50, AMD_UDMA_133 },
 	{ 0 }
 };
 
@@ -489,6 +490,7 @@ static ide_pci_device_t amd74xx_chipsets[] __devinitdata = {
 	/* 13 */ DECLARE_NV_DEV("NFORCE-CK804"),
 	/* 14 */ DECLARE_NV_DEV("NFORCE-MCP04"),
 	/* 15 */ DECLARE_NV_DEV("NFORCE-MCP51"),
+	/* 16 */ DECLARE_NV_DEV("NFORCE-MCP55"),
 };
 
 static int __devinit amd74xx_probe(struct pci_dev *dev, const struct pci_device_id *id)
@@ -524,6 +526,7 @@ static struct pci_device_id amd74xx_pci_tbl[] = {
 	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_CK804_IDE,	PCI_ANY_ID, PCI_ANY_ID, 0, 0, 13 },
 	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP04_IDE,	PCI_ANY_ID, PCI_ANY_ID, 0, 0, 14 },
 	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_IDE,	PCI_ANY_ID, PCI_ANY_ID, 0, 0, 15 },
+	{ PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_IDE,	PCI_ANY_ID, PCI_ANY_ID, 0, 0, 16 },
 	{ 0, },
 };
 MODULE_DEVICE_TABLE(pci, amd74xx_pci_tbl);
diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h
index c3ee1ae4545a..27348c22dacb 100644
--- a/include/linux/pci_ids.h
+++ b/include/linux/pci_ids.h
@@ -1238,6 +1238,7 @@
 #define PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_IDE	0x0265
 #define PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_SATA	0x0266
 #define PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_SATA2	0x0267
+#define PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_IDE	0x036E
 #define PCI_DEVICE_ID_NVIDIA_NVENET_12		0x0268
 #define PCI_DEVICE_ID_NVIDIA_NVENET_13		0x0269
 #define PCI_DEVICE_ID_NVIDIA_MCP51_AUDIO	0x026B
-- 
cgit v1.2.3


From e7270dec080002d8aa18256c756af6c32331ef48 Mon Sep 17 00:00:00 2001
From: Raphael Assenat <raph@raphnet.net>
Date: Mon, 4 Jul 2005 13:23:45 -0700
Subject: [SPARC64/COMPAT]: Add some compat ioctl for ppdev

The following patch adds some ioctls to include/linux/compat_ioctl.h
to allow using ppdev from the 32 bit user space on sparc64.

This patch also adds the PPDEV option in the sparc64 menu, near Parallel
printer support in the 'General machine setup' submenu.

All those ioctls seem to be compatible, since (correct me if I'm wrong)
they dont use the 'long' type. See include/linux/ppdev.h.

The application I used to test the new ioctls only used the following:
PPEXCL
PPCLAIM
PPNEGOT
PPGETMODES
PPRCONTROL
PPWCONTROL
PPDATADIR
PPWDATA
PPRDATA

But I beleive that the other ioctls will work fine.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 arch/sparc64/Kconfig         | 18 ++++++++++++++++++
 include/linux/compat_ioctl.h | 19 ++++++++++++++++++-
 2 files changed, 36 insertions(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/arch/sparc64/Kconfig b/arch/sparc64/Kconfig
index e2b050eb3b96..d78bc13ebbb9 100644
--- a/arch/sparc64/Kconfig
+++ b/arch/sparc64/Kconfig
@@ -444,6 +444,24 @@ config PRINTER
 	  If you have more than 8 printers, you need to increase the LP_NO
 	  macro in lp.c and the PARPORT_MAX macro in parport.h.
 
+config PPDEV
+	tristate "Support for user-space parallel port device drivers"
+	depends on PARPORT
+	---help---
+	  Saying Y to this adds support for /dev/parport device nodes.  This
+	  is needed for programs that want portable access to the parallel
+	  port, for instance deviceid (which displays Plug-and-Play device
+	  IDs).
+
+	  This is the parallel port equivalent of SCSI generic support (sg).
+	  It is safe to say N to this -- it is not needed for normal printing
+	  or parallel port CD-ROM/disk support.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called ppdev.
+
+	  If unsure, say N.
+
 config ENVCTRL
 	tristate "SUNW, envctrl support"
 	depends on PCI
diff --git a/include/linux/compat_ioctl.h b/include/linux/compat_ioctl.h
index 70a4ebb5d964..ecb0d39c0798 100644
--- a/include/linux/compat_ioctl.h
+++ b/include/linux/compat_ioctl.h
@@ -346,10 +346,27 @@ COMPATIBLE_IOCTL(PPPOEIOCDFWD)
 /* LP */
 COMPATIBLE_IOCTL(LPGETSTATUS)
 /* ppdev */
+COMPATIBLE_IOCTL(PPSETMODE)
+COMPATIBLE_IOCTL(PPRSTATUS)
+COMPATIBLE_IOCTL(PPRCONTROL)
+COMPATIBLE_IOCTL(PPWCONTROL)
+COMPATIBLE_IOCTL(PPFCONTROL)
+COMPATIBLE_IOCTL(PPRDATA)
+COMPATIBLE_IOCTL(PPWDATA)
 COMPATIBLE_IOCTL(PPCLAIM)
 COMPATIBLE_IOCTL(PPRELEASE)
-COMPATIBLE_IOCTL(PPEXCL)
 COMPATIBLE_IOCTL(PPYIELD)
+COMPATIBLE_IOCTL(PPEXCL)
+COMPATIBLE_IOCTL(PPDATADIR)
+COMPATIBLE_IOCTL(PPNEGOT)
+COMPATIBLE_IOCTL(PPWCTLONIRQ)
+COMPATIBLE_IOCTL(PPCLRIRQ)
+COMPATIBLE_IOCTL(PPSETPHASE)
+COMPATIBLE_IOCTL(PPGETMODES)
+COMPATIBLE_IOCTL(PPGETMODE)
+COMPATIBLE_IOCTL(PPGETPHASE)
+COMPATIBLE_IOCTL(PPGETFLAGS)
+COMPATIBLE_IOCTL(PPSETFLAGS)
 /* CDROM stuff */
 COMPATIBLE_IOCTL(CDROMPAUSE)
 COMPATIBLE_IOCTL(CDROMRESUME)
-- 
cgit v1.2.3


From 55820ee2f8c767a2833b21bd365e5753f50bd8ce Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 5 Jul 2005 14:08:10 -0700
Subject: [NET]: Fix signedness issues in net/core/filter.c
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

This is the code to load packet data into a register:

                        k = fentry->k;
                        if (k < 0) {
...
                        } else {
                                u32 _tmp, *p;
                                p = skb_header_pointer(skb, k, 4, &_tmp);
                                if (p != NULL) {
                                        A = ntohl(*p);
                                        continue;
                                }
                        }

skb_header_pointer checks if the requested data is within the
linear area:

        int hlen = skb_headlen(skb);

        if (offset + len <= hlen)
                return skb->data + offset;

When offset is within [INT_MAX-len+1..INT_MAX] the addition will
result in a negative number which is <= hlen.

I couldn't trigger a crash on my AMD64 with 2GB of memory, but a
coworker tried on his x86 machine and it crashed immediately.

This patch fixes the check in skb_header_pointer to handle large
positive offsets similar to skb_copy_bits. Invalid data can still
be accessed using negative offsets (also similar to skb_copy_bits),
anyone using negative offsets needs to verify them himself.

Thanks to Thomas Vögtle <thomas.voegtle@coreworks.de> for verifying the
problem by crashing his machine and providing me with an Oops.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/skbuff.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 416a2e4024b2..fbcb18651970 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -1211,7 +1211,7 @@ static inline void *skb_header_pointer(const struct sk_buff *skb, int offset,
 {
 	int hlen = skb_headlen(skb);
 
-	if (offset + len <= hlen)
+	if (hlen - offset >= len)
 		return skb->data + offset;
 
 	if (skb_copy_bits(skb, offset, buffer, len) < 0)
-- 
cgit v1.2.3


From e176fe8954a5239c24afe79b1001ba3c29511963 Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Tue, 5 Jul 2005 14:12:44 -0700
Subject: [NET]: Remove unused security member in sk_buff

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/skbuff.h               | 4 +---
 include/linux/tc_ematch/tc_em_meta.h | 2 +-
 net/core/skbuff.c                    | 2 --
 net/ipv4/ip_output.c                 | 1 -
 net/ipv6/ip6_output.c                | 1 -
 net/sched/em_meta.c                  | 6 ------
 6 files changed, 2 insertions(+), 14 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index fbcb18651970..1e6290f4f81e 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -183,7 +183,6 @@ struct skb_shared_info {
  *	@priority: Packet queueing priority
  *	@users: User count - see {datagram,tcp}.c
  *	@protocol: Packet protocol from driver
- *	@security: Security level of packet
  *	@truesize: Buffer size 
  *	@head: Head of buffer
  *	@data: Data head pointer
@@ -255,8 +254,7 @@ struct sk_buff {
 				pkt_type,
 				ip_summed;
 	__u32			priority;
-	unsigned short		protocol,
-				security;
+	unsigned short		protocol;
 
 	void			(*destructor)(struct sk_buff *skb);
 #ifdef CONFIG_NETFILTER
diff --git a/include/linux/tc_ematch/tc_em_meta.h b/include/linux/tc_ematch/tc_em_meta.h
index a6b2cc530af5..bcb762d93123 100644
--- a/include/linux/tc_ematch/tc_em_meta.h
+++ b/include/linux/tc_ematch/tc_em_meta.h
@@ -45,7 +45,7 @@ enum
 	TCF_META_ID_REALDEV,
 	TCF_META_ID_PRIORITY,
 	TCF_META_ID_PROTOCOL,
-	TCF_META_ID_SECURITY,
+	TCF_META_ID_SECURITY, /* obsolete */
 	TCF_META_ID_PKTTYPE,
 	TCF_META_ID_PKTLEN,
 	TCF_META_ID_DATALEN,
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index bb73b2190ec7..733deee24b9f 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -357,7 +357,6 @@ struct sk_buff *skb_clone(struct sk_buff *skb, int gfp_mask)
 	C(ip_summed);
 	C(priority);
 	C(protocol);
-	C(security);
 	n->destructor = NULL;
 #ifdef CONFIG_NETFILTER
 	C(nfmark);
@@ -422,7 +421,6 @@ static void copy_skb_header(struct sk_buff *new, const struct sk_buff *old)
 	new->pkt_type	= old->pkt_type;
 	new->stamp	= old->stamp;
 	new->destructor = NULL;
-	new->security	= old->security;
 #ifdef CONFIG_NETFILTER
 	new->nfmark	= old->nfmark;
 	new->nfcache	= old->nfcache;
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index 6ce5c3292f9f..1bfa49eda96f 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -389,7 +389,6 @@ static void ip_copy_metadata(struct sk_buff *to, struct sk_buff *from)
 	to->pkt_type = from->pkt_type;
 	to->priority = from->priority;
 	to->protocol = from->protocol;
-	to->security = from->security;
 	dst_release(to->dst);
 	to->dst = dst_clone(from->dst);
 	to->dev = from->dev;
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index 06e7cdaeedc5..1f2c2f9e353f 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -465,7 +465,6 @@ static void ip6_copy_metadata(struct sk_buff *to, struct sk_buff *from)
 	to->pkt_type = from->pkt_type;
 	to->priority = from->priority;
 	to->protocol = from->protocol;
-	to->security = from->security;
 	dst_release(to->dst);
 	to->dst = dst_clone(from->dst);
 	to->dev = from->dev;
diff --git a/net/sched/em_meta.c b/net/sched/em_meta.c
index 48bb23c2a35a..53d98f8d3d80 100644
--- a/net/sched/em_meta.c
+++ b/net/sched/em_meta.c
@@ -205,11 +205,6 @@ META_COLLECTOR(int_protocol)
 	dst->value = skb->protocol;
 }
 
-META_COLLECTOR(int_security)
-{
-	dst->value = skb->security;
-}
-
 META_COLLECTOR(int_pkttype)
 {
 	dst->value = skb->pkt_type;
@@ -524,7 +519,6 @@ static struct meta_ops __meta_ops[TCF_META_TYPE_MAX+1][TCF_META_ID_MAX+1] = {
 		[META_ID(REALDEV)]		= META_FUNC(int_realdev),
 		[META_ID(PRIORITY)]		= META_FUNC(int_priority),
 		[META_ID(PROTOCOL)]		= META_FUNC(int_protocol),
-		[META_ID(SECURITY)]		= META_FUNC(int_security),
 		[META_ID(PKTTYPE)]		= META_FUNC(int_pkttype),
 		[META_ID(PKTLEN)]		= META_FUNC(int_pktlen),
 		[META_ID(DATALEN)]		= META_FUNC(int_datalen),
-- 
cgit v1.2.3


From 1cbb3380ef683f742876f48e3739b3df4ea9e168 Mon Sep 17 00:00:00 2001
From: Thomas Graf <tgraf@suug.ch>
Date: Tue, 5 Jul 2005 14:13:41 -0700
Subject: [NET]: Reduce size of sk_buff by 4 bytes

Reduce local_df to a bit field and ip_summed to a 2 bits
field thus saving 13 bits. Move bit fields, packet type,
and protocol into the spare area between the priority
and the destructor. Saves 4 bytes on both, 32bit and
64bit architectures.

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/skbuff.h | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 1e6290f4f81e..14b950413495 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -248,17 +248,18 @@ struct sk_buff {
 				data_len,
 				mac_len,
 				csum;
-	unsigned char		local_df,
-				cloned:1,
-				nohdr:1,
-				pkt_type,
-				ip_summed;
 	__u32			priority;
-	unsigned short		protocol;
+	__u8			local_df:1,
+				cloned:1,
+				ip_summed:2,
+				nohdr:1;
+				/* 3 bits spare */
+	__u8			pkt_type;
+	__u16			protocol;
 
 	void			(*destructor)(struct sk_buff *skb);
 #ifdef CONFIG_NETFILTER
-        unsigned long		nfmark;
+	unsigned long		nfmark;
 	__u32			nfcache;
 	__u32			nfctinfo;
 	struct nf_conntrack	*nfct;
-- 
cgit v1.2.3


From bc971dee6ece1fd0d431948924becd9c50e7b778 Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@lst.de>
Date: Tue, 5 Jul 2005 15:03:46 -0700
Subject: [SHAPER]: Switch to spinlocks.

Dave, you were right and the sleeping locks in shaper were
broken. Markus Kanet noticed this and also tested the patch below that
switches locking to spinlocks.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/shaper.c      | 42 ++++++++++++++++--------------------------
 include/linux/if_shaper.h |  2 +-
 2 files changed, 17 insertions(+), 27 deletions(-)

(limited to 'include/linux')

diff --git a/drivers/net/shaper.c b/drivers/net/shaper.c
index 20edeb345792..3ad0b6751f6f 100644
--- a/drivers/net/shaper.c
+++ b/drivers/net/shaper.c
@@ -135,10 +135,8 @@ static int shaper_start_xmit(struct sk_buff *skb, struct net_device *dev)
 {
 	struct shaper *shaper = dev->priv;
  	struct sk_buff *ptr;
-   
-	if (down_trylock(&shaper->sem))
-		return -1;
-
+  
+	spin_lock(&shaper->lock);
  	ptr=shaper->sendq.prev;
  	
  	/*
@@ -232,7 +230,7 @@ static int shaper_start_xmit(struct sk_buff *skb, struct net_device *dev)
                 shaper->stats.collisions++;
  	}
 	shaper_kick(shaper);
-	up(&shaper->sem);
+	spin_unlock(&shaper->lock);
  	return 0;
 }
 
@@ -271,11 +269,9 @@ static void shaper_timer(unsigned long data)
 {
 	struct shaper *shaper = (struct shaper *)data;
 
-	if (!down_trylock(&shaper->sem)) {
-		shaper_kick(shaper);
-		up(&shaper->sem);
-	} else
-		mod_timer(&shaper->timer, jiffies);
+	spin_lock(&shaper->lock);
+	shaper_kick(shaper);
+	spin_unlock(&shaper->lock);
 }
 
 /*
@@ -331,21 +327,6 @@ static void shaper_kick(struct shaper *shaper)
 }
 
 
-/*
- *	Flush the shaper queues on a closedown
- */
- 
-static void shaper_flush(struct shaper *shaper)
-{
-	struct sk_buff *skb;
-
-	down(&shaper->sem);
-	while((skb=skb_dequeue(&shaper->sendq))!=NULL)
-		dev_kfree_skb(skb);
-	shaper_kick(shaper);
-	up(&shaper->sem);
-}
-
 /*
  *	Bring the interface up. We just disallow this until a 
  *	bind.
@@ -375,7 +356,15 @@ static int shaper_open(struct net_device *dev)
 static int shaper_close(struct net_device *dev)
 {
 	struct shaper *shaper=dev->priv;
-	shaper_flush(shaper);
+	struct sk_buff *skb;
+
+	while ((skb = skb_dequeue(&shaper->sendq)) != NULL)
+		dev_kfree_skb(skb);
+
+	spin_lock_bh(&shaper->lock);
+	shaper_kick(shaper);
+	spin_unlock_bh(&shaper->lock);
+
 	del_timer_sync(&shaper->timer);
 	return 0;
 }
@@ -576,6 +565,7 @@ static void shaper_init_priv(struct net_device *dev)
 	init_timer(&sh->timer);
 	sh->timer.function=shaper_timer;
 	sh->timer.data=(unsigned long)sh;
+	spin_lock_init(&sh->lock);
 }
 
 /*
diff --git a/include/linux/if_shaper.h b/include/linux/if_shaper.h
index 004e6f09a6e2..68c896a36a34 100644
--- a/include/linux/if_shaper.h
+++ b/include/linux/if_shaper.h
@@ -23,7 +23,7 @@ struct shaper
 	__u32 shapeclock;
 	unsigned long recovery;	/* Time we can next clock a packet out on
 				   an empty queue */
-	struct semaphore sem;
+	spinlock_t lock;
         struct net_device_stats stats;
 	struct net_device *dev;
 	int  (*hard_start_xmit) (struct sk_buff *skb,
-- 
cgit v1.2.3


From c1b4a7e69576d65efc31a8cea0714173c2841244 Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@davemloft.net>
Date: Tue, 5 Jul 2005 15:24:38 -0700
Subject: [TCP]: Move to new TSO segmenting scheme.

Make TSO segment transmit size decisions at send time not earlier.

The basic scheme is that we try to build as large a TSO frame as
possible when pulling in the user data, but the size of the TSO frame
output to the card is determined at transmit time.

This is guided by tp->xmit_size_goal.  It is always set to a multiple
of MSS and tells sendmsg/sendpage how large an SKB to try and build.

Later, tcp_write_xmit() and tcp_push_one() chop up the packet if
necessary and conditions warrant.  These routines can also decide to
"defer" in order to wait for more ACKs to arrive and thus allow larger
TSO frames to be emitted.

A general observation is that TSO elongates the pipe, thus requiring a
larger congestion window and larger buffering especially at the sender
side.  Therefore, it is important that applications 1) get a large
enough socket send buffer (this is accomplished by our dynamic send
buffer expansion code) 2) do large enough writes.

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/tcp.h   |   2 +-
 include/net/tcp.h     |   4 +-
 net/ipv4/tcp.c        |  26 ++-
 net/ipv4/tcp_input.c  |  10 +-
 net/ipv4/tcp_ipv4.c   |   2 +-
 net/ipv4/tcp_output.c | 578 +++++++++++++++++++++++++++++++-------------------
 net/ipv6/tcp_ipv6.c   |   2 +-
 7 files changed, 384 insertions(+), 240 deletions(-)

(limited to 'include/linux')

diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index dfd93d03f5d2..e4fd82e42104 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -286,7 +286,7 @@ struct tcp_sock {
 	__u32	max_window;	/* Maximal window ever seen from peer	*/
 	__u32	pmtu_cookie;	/* Last pmtu seen by socket		*/
 	__u32	mss_cache;	/* Cached effective mss, not including SACKS */
-	__u16	mss_cache_std;	/* Like mss_cache, but without TSO */
+	__u16	xmit_size_goal;	/* Goal for segmenting output packets	*/
 	__u16	ext_header_len;	/* Network protocol overhead (IP/IPv6 options) */
 	__u8	ca_state;	/* State of fast-retransmit machine 	*/
 	__u8	retransmits;	/* Number of unrecovered RTO timeouts.	*/
diff --git a/include/net/tcp.h b/include/net/tcp.h
index b19238027da8..a166918ca56d 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -862,7 +862,7 @@ extern int  tcp_write_wakeup(struct sock *);
 extern void tcp_send_fin(struct sock *sk);
 extern void tcp_send_active_reset(struct sock *sk, int priority);
 extern int  tcp_send_synack(struct sock *);
-extern void tcp_push_one(struct sock *, unsigned mss_now);
+extern void tcp_push_one(struct sock *, unsigned int mss_now);
 extern void tcp_send_ack(struct sock *sk);
 extern void tcp_send_delayed_ack(struct sock *sk);
 
@@ -968,7 +968,7 @@ static inline void tcp_reset_xmit_timer(struct sock *sk, int what, unsigned long
 static inline void tcp_initialize_rcv_mss(struct sock *sk)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
-	unsigned int hint = min(tp->advmss, tp->mss_cache_std);
+	unsigned int hint = min_t(unsigned int, tp->advmss, tp->mss_cache);
 
 	hint = min(hint, tp->rcv_wnd/2);
 	hint = min(hint, TCP_MIN_RCVMSS);
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 2ba73bf3a8f9..29894c749163 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -615,7 +615,7 @@ static ssize_t do_tcp_sendpages(struct sock *sk, struct page **pages, int poffse
 			 size_t psize, int flags)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
-	int mss_now;
+	int mss_now, size_goal;
 	int err;
 	ssize_t copied;
 	long timeo = sock_sndtimeo(sk, flags & MSG_DONTWAIT);
@@ -628,6 +628,7 @@ static ssize_t do_tcp_sendpages(struct sock *sk, struct page **pages, int poffse
 	clear_bit(SOCK_ASYNC_NOSPACE, &sk->sk_socket->flags);
 
 	mss_now = tcp_current_mss(sk, !(flags&MSG_OOB));
+	size_goal = tp->xmit_size_goal;
 	copied = 0;
 
 	err = -EPIPE;
@@ -641,7 +642,7 @@ static ssize_t do_tcp_sendpages(struct sock *sk, struct page **pages, int poffse
 		int offset = poffset % PAGE_SIZE;
 		int size = min_t(size_t, psize, PAGE_SIZE - offset);
 
-		if (!sk->sk_send_head || (copy = mss_now - skb->len) <= 0) {
+		if (!sk->sk_send_head || (copy = size_goal - skb->len) <= 0) {
 new_segment:
 			if (!sk_stream_memory_free(sk))
 				goto wait_for_sndbuf;
@@ -652,7 +653,7 @@ new_segment:
 				goto wait_for_memory;
 
 			skb_entail(sk, tp, skb);
-			copy = mss_now;
+			copy = size_goal;
 		}
 
 		if (copy > size)
@@ -693,7 +694,7 @@ new_segment:
 		if (!(psize -= copy))
 			goto out;
 
-		if (skb->len != mss_now || (flags & MSG_OOB))
+		if (skb->len < mss_now || (flags & MSG_OOB))
 			continue;
 
 		if (forced_push(tp)) {
@@ -713,6 +714,7 @@ wait_for_memory:
 			goto do_error;
 
 		mss_now = tcp_current_mss(sk, !(flags&MSG_OOB));
+		size_goal = tp->xmit_size_goal;
 	}
 
 out:
@@ -754,7 +756,7 @@ ssize_t tcp_sendpage(struct socket *sock, struct page *page, int offset,
 
 static inline int select_size(struct sock *sk, struct tcp_sock *tp)
 {
-	int tmp = tp->mss_cache_std;
+	int tmp = tp->mss_cache;
 
 	if (sk->sk_route_caps & NETIF_F_SG) {
 		if (sk->sk_route_caps & NETIF_F_TSO)
@@ -778,7 +780,7 @@ int tcp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct sk_buff *skb;
 	int iovlen, flags;
-	int mss_now;
+	int mss_now, size_goal;
 	int err, copied;
 	long timeo;
 
@@ -797,6 +799,7 @@ int tcp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
 	clear_bit(SOCK_ASYNC_NOSPACE, &sk->sk_socket->flags);
 
 	mss_now = tcp_current_mss(sk, !(flags&MSG_OOB));
+	size_goal = tp->xmit_size_goal;
 
 	/* Ok commence sending. */
 	iovlen = msg->msg_iovlen;
@@ -819,7 +822,7 @@ int tcp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
 			skb = sk->sk_write_queue.prev;
 
 			if (!sk->sk_send_head ||
-			    (copy = mss_now - skb->len) <= 0) {
+			    (copy = size_goal - skb->len) <= 0) {
 
 new_segment:
 				/* Allocate new segment. If the interface is SG,
@@ -842,7 +845,7 @@ new_segment:
 					skb->ip_summed = CHECKSUM_HW;
 
 				skb_entail(sk, tp, skb);
-				copy = mss_now;
+				copy = size_goal;
 			}
 
 			/* Try to append data to the end of skb. */
@@ -937,7 +940,7 @@ new_segment:
 			if ((seglen -= copy) == 0 && iovlen == 0)
 				goto out;
 
-			if (skb->len != mss_now || (flags & MSG_OOB))
+			if (skb->len < mss_now || (flags & MSG_OOB))
 				continue;
 
 			if (forced_push(tp)) {
@@ -957,6 +960,7 @@ wait_for_memory:
 				goto do_error;
 
 			mss_now = tcp_current_mss(sk, !(flags&MSG_OOB));
+			size_goal = tp->xmit_size_goal;
 		}
 	}
 
@@ -2128,7 +2132,7 @@ void tcp_get_info(struct sock *sk, struct tcp_info *info)
 
 	info->tcpi_rto = jiffies_to_usecs(tp->rto);
 	info->tcpi_ato = jiffies_to_usecs(tp->ack.ato);
-	info->tcpi_snd_mss = tp->mss_cache_std;
+	info->tcpi_snd_mss = tp->mss_cache;
 	info->tcpi_rcv_mss = tp->ack.rcv_mss;
 
 	info->tcpi_unacked = tp->packets_out;
@@ -2178,7 +2182,7 @@ int tcp_getsockopt(struct sock *sk, int level, int optname, char __user *optval,
 
 	switch (optname) {
 	case TCP_MAXSEG:
-		val = tp->mss_cache_std;
+		val = tp->mss_cache;
 		if (!val && ((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_LISTEN)))
 			val = tp->rx_opt.user_mss;
 		break;
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 2ef2f355b8b8..8de2f1071c2b 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -740,10 +740,10 @@ __u32 tcp_init_cwnd(struct tcp_sock *tp, struct dst_entry *dst)
 	__u32 cwnd = (dst ? dst_metric(dst, RTAX_INITCWND) : 0);
 
 	if (!cwnd) {
-		if (tp->mss_cache_std > 1460)
+		if (tp->mss_cache > 1460)
 			cwnd = 2;
 		else
-			cwnd = (tp->mss_cache_std > 1095) ? 3 : 4;
+			cwnd = (tp->mss_cache > 1095) ? 3 : 4;
 	}
 	return min_t(__u32, cwnd, tp->snd_cwnd_clamp);
 }
@@ -914,7 +914,7 @@ tcp_sacktag_write_queue(struct sock *sk, struct sk_buff *ack_skb, u32 prior_snd_
 	if (sk->sk_route_caps & NETIF_F_TSO) {
 		sk->sk_route_caps &= ~NETIF_F_TSO;
 		sock_set_flag(sk, SOCK_NO_LARGESEND);
-		tp->mss_cache = tp->mss_cache_std;
+		tp->mss_cache = tp->mss_cache;
 	}
 
 	if (!tp->sacked_out)
@@ -1077,7 +1077,7 @@ tcp_sacktag_write_queue(struct sock *sk, struct sk_buff *ack_skb, u32 prior_snd_
 			    (IsFack(tp) ||
 			     !before(lost_retrans,
 				     TCP_SKB_CB(skb)->ack_seq + tp->reordering *
-				     tp->mss_cache_std))) {
+				     tp->mss_cache))) {
 				TCP_SKB_CB(skb)->sacked &= ~TCPCB_SACKED_RETRANS;
 				tp->retrans_out -= tcp_skb_pcount(skb);
 
@@ -3334,7 +3334,7 @@ static void tcp_new_space(struct sock *sk)
 	struct tcp_sock *tp = tcp_sk(sk);
 
 	if (tcp_should_expand_sndbuf(sk, tp)) {
- 		int sndmem = max_t(u32, tp->rx_opt.mss_clamp, tp->mss_cache_std) +
+ 		int sndmem = max_t(u32, tp->rx_opt.mss_clamp, tp->mss_cache) +
 			MAX_TCP_HEADER + 16 + sizeof(struct sk_buff),
 		    demanded = max_t(unsigned int, tp->snd_cwnd,
 						   tp->reordering + 1);
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index ebf112347a97..62f62bb05c2a 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -2045,7 +2045,7 @@ static int tcp_v4_init_sock(struct sock *sk)
 	 */
 	tp->snd_ssthresh = 0x7fffffff;	/* Infinity */
 	tp->snd_cwnd_clamp = ~0;
-	tp->mss_cache_std = tp->mss_cache = 536;
+	tp->mss_cache = 536;
 
 	tp->reordering = sysctl_tcp_reordering;
 	tp->ca_ops = &tcp_init_congestion_ops;
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 0a4cd24b6578..fd3ce38184ae 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -49,7 +49,7 @@ int sysctl_tcp_retrans_collapse = 1;
  * will allow a single TSO frame to consume.  Building TSO frames
  * which are too large can cause TCP streams to be bursty.
  */
-int sysctl_tcp_tso_win_divisor = 8;
+int sysctl_tcp_tso_win_divisor = 3;
 
 static inline void update_send_head(struct sock *sk, struct tcp_sock *tp,
 				    struct sk_buff *skb)
@@ -403,21 +403,11 @@ static void tcp_queue_skb(struct sock *sk, struct sk_buff *skb)
 		sk->sk_send_head = skb;
 }
 
-static inline void tcp_tso_set_push(struct sk_buff *skb)
-{
-	/* Force push to be on for any TSO frames to workaround
-	 * problems with busted implementations like Mac OS-X that
-	 * hold off socket receive wakeups until push is seen.
-	 */
-	if (tcp_skb_pcount(skb) > 1)
-		TCP_SKB_CB(skb)->flags |= TCPCB_FLAG_PSH;
-}
-
 static void tcp_set_skb_tso_segs(struct sock *sk, struct sk_buff *skb)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 
-	if (skb->len <= tp->mss_cache_std ||
+	if (skb->len <= tp->mss_cache ||
 	    !(sk->sk_route_caps & NETIF_F_TSO)) {
 		/* Avoid the costly divide in the normal
 		 * non-TSO case.
@@ -427,164 +417,10 @@ static void tcp_set_skb_tso_segs(struct sock *sk, struct sk_buff *skb)
 	} else {
 		unsigned int factor;
 
-		factor = skb->len + (tp->mss_cache_std - 1);
-		factor /= tp->mss_cache_std;
+		factor = skb->len + (tp->mss_cache - 1);
+		factor /= tp->mss_cache;
 		skb_shinfo(skb)->tso_segs = factor;
-		skb_shinfo(skb)->tso_size = tp->mss_cache_std;
-	}
-}
-
-/* Does SKB fit into the send window? */
-static inline int tcp_snd_wnd_test(struct tcp_sock *tp, struct sk_buff *skb, unsigned int cur_mss)
-{
-	u32 end_seq = TCP_SKB_CB(skb)->end_seq;
-
-	return !after(end_seq, tp->snd_una + tp->snd_wnd);
-}
-
-/* Can at least one segment of SKB be sent right now, according to the
- * congestion window rules?  If so, return how many segments are allowed.
- */
-static inline unsigned int tcp_cwnd_test(struct tcp_sock *tp, struct sk_buff *skb)
-{
-	u32 in_flight, cwnd;
-
-	/* Don't be strict about the congestion window for the final FIN.  */
-	if (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN)
-		return 1;
-
-	in_flight = tcp_packets_in_flight(tp);
-	cwnd = tp->snd_cwnd;
-	if (in_flight < cwnd)
-		return (cwnd - in_flight);
-
-	return 0;
-}
-
-static inline int tcp_minshall_check(const struct tcp_sock *tp)
-{
-	return after(tp->snd_sml,tp->snd_una) &&
-		!after(tp->snd_sml, tp->snd_nxt);
-}
-
-/* Return 0, if packet can be sent now without violation Nagle's rules:
- * 1. It is full sized.
- * 2. Or it contains FIN. (already checked by caller)
- * 3. Or TCP_NODELAY was set.
- * 4. Or TCP_CORK is not set, and all sent packets are ACKed.
- *    With Minshall's modification: all sent small packets are ACKed.
- */
-
-static inline int tcp_nagle_check(const struct tcp_sock *tp,
-				  const struct sk_buff *skb, 
-				  unsigned mss_now, int nonagle)
-{
-	return (skb->len < mss_now &&
-		((nonagle&TCP_NAGLE_CORK) ||
-		 (!nonagle &&
-		  tp->packets_out &&
-		  tcp_minshall_check(tp))));
-}
-
-/* Return non-zero if the Nagle test allows this packet to be
- * sent now.
- */
-static inline int tcp_nagle_test(struct tcp_sock *tp, struct sk_buff *skb,
-				 unsigned int cur_mss, int nonagle)
-{
-	/* Nagle rule does not apply to frames, which sit in the middle of the
-	 * write_queue (they have no chances to get new data).
-	 *
-	 * This is implemented in the callers, where they modify the 'nonagle'
-	 * argument based upon the location of SKB in the send queue.
-	 */
-	if (nonagle & TCP_NAGLE_PUSH)
-		return 1;
-
-	/* Don't use the nagle rule for urgent data (or for the final FIN).  */
-	if (tp->urg_mode ||
-	    (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN))
-		return 1;
-
-	if (!tcp_nagle_check(tp, skb, cur_mss, nonagle))
-		return 1;
-
-	return 0;
-}
-
-/* This must be invoked the first time we consider transmitting
- * SKB onto the wire.
- */
-static inline int tcp_init_tso_segs(struct sock *sk, struct sk_buff *skb)
-{
-	int tso_segs = tcp_skb_pcount(skb);
-
-	if (!tso_segs) {
-		tcp_set_skb_tso_segs(sk, skb);
-		tso_segs = tcp_skb_pcount(skb);
-	}
-	return tso_segs;
-}
-
-/* This checks if the data bearing packet SKB (usually sk->sk_send_head)
- * should be put on the wire right now.  If so, it returns the number of
- * packets allowed by the congestion window.
- */
-static unsigned int tcp_snd_test(struct sock *sk, struct sk_buff *skb,
-				 unsigned int cur_mss, int nonagle)
-{
-	struct tcp_sock *tp = tcp_sk(sk);
-	unsigned int cwnd_quota;
-
-	tcp_init_tso_segs(sk, skb);
-
-	if (!tcp_nagle_test(tp, skb, cur_mss, nonagle))
-		return 0;
-
-	cwnd_quota = tcp_cwnd_test(tp, skb);
-	if (cwnd_quota &&
-	    !tcp_snd_wnd_test(tp, skb, cur_mss))
-		cwnd_quota = 0;
-
-	return cwnd_quota;
-}
-
-static inline int tcp_skb_is_last(const struct sock *sk, 
-				  const struct sk_buff *skb)
-{
-	return skb->next == (struct sk_buff *)&sk->sk_write_queue;
-}
-
-int tcp_may_send_now(struct sock *sk, struct tcp_sock *tp)
-{
-	struct sk_buff *skb = sk->sk_send_head;
-
-	return (skb &&
-		tcp_snd_test(sk, skb, tcp_current_mss(sk, 1),
-			     (tcp_skb_is_last(sk, skb) ?
-			      TCP_NAGLE_PUSH :
-			      tp->nonagle)));
-}
-
-
-/* Send _single_ skb sitting at the send head. This function requires
- * true push pending frames to setup probe timer etc.
- */
-void tcp_push_one(struct sock *sk, unsigned cur_mss)
-{
-	struct tcp_sock *tp = tcp_sk(sk);
-	struct sk_buff *skb = sk->sk_send_head;
-
-	if (tcp_snd_test(sk, skb, cur_mss, TCP_NAGLE_PUSH)) {
-		/* Send it out now. */
-		TCP_SKB_CB(skb)->when = tcp_time_stamp;
-		tcp_tso_set_push(skb);
-		if (!tcp_transmit_skb(sk, skb_clone(skb, sk->sk_allocation))) {
-			sk->sk_send_head = NULL;
-			tp->snd_nxt = TCP_SKB_CB(skb)->end_seq;
-			tcp_packets_out_inc(sk, tp, skb);
-			return;
-		}
+		skb_shinfo(skb)->tso_size = tp->mss_cache;
 	}
 }
 
@@ -791,7 +627,7 @@ unsigned int tcp_sync_mss(struct sock *sk, u32 pmtu)
 
 	/* And store cached results */
 	tp->pmtu_cookie = pmtu;
-	tp->mss_cache = tp->mss_cache_std = mss_now;
+	tp->mss_cache = mss_now;
 
 	return mss_now;
 }
@@ -803,56 +639,47 @@ unsigned int tcp_sync_mss(struct sock *sk, u32 pmtu)
  * cannot be large. However, taking into account rare use of URG, this
  * is not a big flaw.
  */
-
-unsigned int tcp_current_mss(struct sock *sk, int large)
+unsigned int tcp_current_mss(struct sock *sk, int large_allowed)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct dst_entry *dst = __sk_dst_get(sk);
-	unsigned int do_large, mss_now;
+	u32 mss_now;
+	u16 xmit_size_goal;
+	int doing_tso = 0;
+
+	mss_now = tp->mss_cache;
+
+	if (large_allowed &&
+	    (sk->sk_route_caps & NETIF_F_TSO) &&
+	    !tp->urg_mode)
+		doing_tso = 1;
 
-	mss_now = tp->mss_cache_std;
 	if (dst) {
 		u32 mtu = dst_mtu(dst);
 		if (mtu != tp->pmtu_cookie)
 			mss_now = tcp_sync_mss(sk, mtu);
 	}
 
-	do_large = (large &&
-		    (sk->sk_route_caps & NETIF_F_TSO) &&
-		    !tp->urg_mode);
+	if (tp->rx_opt.eff_sacks)
+		mss_now -= (TCPOLEN_SACK_BASE_ALIGNED +
+			    (tp->rx_opt.eff_sacks * TCPOLEN_SACK_PERBLOCK));
 
-	if (do_large) {
-		unsigned int large_mss, factor, limit;
+	xmit_size_goal = mss_now;
 
-		large_mss = 65535 - tp->af_specific->net_header_len -
+	if (doing_tso) {
+		xmit_size_goal = 65535 -
+			tp->af_specific->net_header_len -
 			tp->ext_header_len - tp->tcp_header_len;
 
-		if (tp->max_window && large_mss > (tp->max_window>>1))
-			large_mss = max((tp->max_window>>1),
-					68U - tp->tcp_header_len);
-
-		factor = large_mss / mss_now;
+		if (tp->max_window &&
+		    (xmit_size_goal > (tp->max_window >> 1)))
+			xmit_size_goal = max((tp->max_window >> 1),
+					     68U - tp->tcp_header_len);
 
-		/* Always keep large mss multiple of real mss, but
-		 * do not exceed 1/tso_win_divisor of the congestion window
-		 * so we can keep the ACK clock ticking and minimize
-		 * bursting.
-		 */
-		limit = tp->snd_cwnd;
-		if (sysctl_tcp_tso_win_divisor)
-			limit /= sysctl_tcp_tso_win_divisor;
-		limit = max(1U, limit);
-		if (factor > limit)
-			factor = limit;
-
-		tp->mss_cache = mss_now * factor;
-
-		mss_now = tp->mss_cache;
+		xmit_size_goal -= (xmit_size_goal % mss_now);
 	}
+	tp->xmit_size_goal = xmit_size_goal;
 
-	if (tp->rx_opt.eff_sacks)
-		mss_now -= (TCPOLEN_SACK_BASE_ALIGNED +
-			    (tp->rx_opt.eff_sacks * TCPOLEN_SACK_PERBLOCK));
 	return mss_now;
 }
 
@@ -876,6 +703,251 @@ static inline void tcp_cwnd_validate(struct sock *sk, struct tcp_sock *tp)
 	}
 }
 
+static unsigned int tcp_window_allows(struct tcp_sock *tp, struct sk_buff *skb, unsigned int mss_now, unsigned int cwnd)
+{
+	u32 window, cwnd_len;
+
+	window = (tp->snd_una + tp->snd_wnd - TCP_SKB_CB(skb)->seq);
+	cwnd_len = mss_now * cwnd;
+	return min(window, cwnd_len);
+}
+
+/* Can at least one segment of SKB be sent right now, according to the
+ * congestion window rules?  If so, return how many segments are allowed.
+ */
+static inline unsigned int tcp_cwnd_test(struct tcp_sock *tp, struct sk_buff *skb)
+{
+	u32 in_flight, cwnd;
+
+	/* Don't be strict about the congestion window for the final FIN.  */
+	if (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN)
+		return 1;
+
+	in_flight = tcp_packets_in_flight(tp);
+	cwnd = tp->snd_cwnd;
+	if (in_flight < cwnd)
+		return (cwnd - in_flight);
+
+	return 0;
+}
+
+/* This must be invoked the first time we consider transmitting
+ * SKB onto the wire.
+ */
+static inline int tcp_init_tso_segs(struct sock *sk, struct sk_buff *skb)
+{
+	int tso_segs = tcp_skb_pcount(skb);
+
+	if (!tso_segs) {
+		tcp_set_skb_tso_segs(sk, skb);
+		tso_segs = tcp_skb_pcount(skb);
+	}
+	return tso_segs;
+}
+
+static inline int tcp_minshall_check(const struct tcp_sock *tp)
+{
+	return after(tp->snd_sml,tp->snd_una) &&
+		!after(tp->snd_sml, tp->snd_nxt);
+}
+
+/* Return 0, if packet can be sent now without violation Nagle's rules:
+ * 1. It is full sized.
+ * 2. Or it contains FIN. (already checked by caller)
+ * 3. Or TCP_NODELAY was set.
+ * 4. Or TCP_CORK is not set, and all sent packets are ACKed.
+ *    With Minshall's modification: all sent small packets are ACKed.
+ */
+
+static inline int tcp_nagle_check(const struct tcp_sock *tp,
+				  const struct sk_buff *skb, 
+				  unsigned mss_now, int nonagle)
+{
+	return (skb->len < mss_now &&
+		((nonagle&TCP_NAGLE_CORK) ||
+		 (!nonagle &&
+		  tp->packets_out &&
+		  tcp_minshall_check(tp))));
+}
+
+/* Return non-zero if the Nagle test allows this packet to be
+ * sent now.
+ */
+static inline int tcp_nagle_test(struct tcp_sock *tp, struct sk_buff *skb,
+				 unsigned int cur_mss, int nonagle)
+{
+	/* Nagle rule does not apply to frames, which sit in the middle of the
+	 * write_queue (they have no chances to get new data).
+	 *
+	 * This is implemented in the callers, where they modify the 'nonagle'
+	 * argument based upon the location of SKB in the send queue.
+	 */
+	if (nonagle & TCP_NAGLE_PUSH)
+		return 1;
+
+	/* Don't use the nagle rule for urgent data (or for the final FIN).  */
+	if (tp->urg_mode ||
+	    (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN))
+		return 1;
+
+	if (!tcp_nagle_check(tp, skb, cur_mss, nonagle))
+		return 1;
+
+	return 0;
+}
+
+/* Does at least the first segment of SKB fit into the send window? */
+static inline int tcp_snd_wnd_test(struct tcp_sock *tp, struct sk_buff *skb, unsigned int cur_mss)
+{
+	u32 end_seq = TCP_SKB_CB(skb)->end_seq;
+
+	if (skb->len > cur_mss)
+		end_seq = TCP_SKB_CB(skb)->seq + cur_mss;
+
+	return !after(end_seq, tp->snd_una + tp->snd_wnd);
+}
+
+/* This checks if the data bearing packet SKB (usually sk->sk_send_head)
+ * should be put on the wire right now.  If so, it returns the number of
+ * packets allowed by the congestion window.
+ */
+static unsigned int tcp_snd_test(struct sock *sk, struct sk_buff *skb,
+				 unsigned int cur_mss, int nonagle)
+{
+	struct tcp_sock *tp = tcp_sk(sk);
+	unsigned int cwnd_quota;
+
+	tcp_init_tso_segs(sk, skb);
+
+	if (!tcp_nagle_test(tp, skb, cur_mss, nonagle))
+		return 0;
+
+	cwnd_quota = tcp_cwnd_test(tp, skb);
+	if (cwnd_quota &&
+	    !tcp_snd_wnd_test(tp, skb, cur_mss))
+		cwnd_quota = 0;
+
+	return cwnd_quota;
+}
+
+static inline int tcp_skb_is_last(const struct sock *sk, 
+				  const struct sk_buff *skb)
+{
+	return skb->next == (struct sk_buff *)&sk->sk_write_queue;
+}
+
+int tcp_may_send_now(struct sock *sk, struct tcp_sock *tp)
+{
+	struct sk_buff *skb = sk->sk_send_head;
+
+	return (skb &&
+		tcp_snd_test(sk, skb, tcp_current_mss(sk, 1),
+			     (tcp_skb_is_last(sk, skb) ?
+			      TCP_NAGLE_PUSH :
+			      tp->nonagle)));
+}
+
+/* Trim TSO SKB to LEN bytes, put the remaining data into a new packet
+ * which is put after SKB on the list.  It is very much like
+ * tcp_fragment() except that it may make several kinds of assumptions
+ * in order to speed up the splitting operation.  In particular, we
+ * know that all the data is in scatter-gather pages, and that the
+ * packet has never been sent out before (and thus is not cloned).
+ */
+static int tso_fragment(struct sock *sk, struct sk_buff *skb, unsigned int len)
+{
+	struct sk_buff *buff;
+	int nlen = skb->len - len;
+	u16 flags;
+
+	/* All of a TSO frame must be composed of paged data.  */
+	BUG_ON(skb->len != skb->data_len);
+
+	buff = sk_stream_alloc_pskb(sk, 0, 0, GFP_ATOMIC);
+	if (unlikely(buff == NULL))
+		return -ENOMEM;
+
+	buff->truesize = nlen;
+	skb->truesize -= nlen;
+
+	/* Correct the sequence numbers. */
+	TCP_SKB_CB(buff)->seq = TCP_SKB_CB(skb)->seq + len;
+	TCP_SKB_CB(buff)->end_seq = TCP_SKB_CB(skb)->end_seq;
+	TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(buff)->seq;
+
+	/* PSH and FIN should only be set in the second packet. */
+	flags = TCP_SKB_CB(skb)->flags;
+	TCP_SKB_CB(skb)->flags = flags & ~(TCPCB_FLAG_FIN|TCPCB_FLAG_PSH);
+	TCP_SKB_CB(buff)->flags = flags;
+
+	/* This packet was never sent out yet, so no SACK bits. */
+	TCP_SKB_CB(buff)->sacked = 0;
+
+	buff->ip_summed = skb->ip_summed = CHECKSUM_HW;
+	skb_split(skb, buff, len);
+
+	/* Fix up tso_factor for both original and new SKB.  */
+	tcp_set_skb_tso_segs(sk, skb);
+	tcp_set_skb_tso_segs(sk, buff);
+
+	/* Link BUFF into the send queue. */
+	skb_header_release(buff);
+	__skb_append(skb, buff);
+
+	return 0;
+}
+
+/* Try to defer sending, if possible, in order to minimize the amount
+ * of TSO splitting we do.  View it as a kind of TSO Nagle test.
+ *
+ * This algorithm is from John Heffner.
+ */
+static int tcp_tso_should_defer(struct sock *sk, struct tcp_sock *tp, struct sk_buff *skb)
+{
+	u32 send_win, cong_win, limit, in_flight;
+
+	if (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN)
+		return 0;
+
+	in_flight = tcp_packets_in_flight(tp);
+
+	BUG_ON(tcp_skb_pcount(skb) <= 1 ||
+	       (tp->snd_cwnd <= in_flight));
+
+	send_win = (tp->snd_una + tp->snd_wnd) - TCP_SKB_CB(skb)->seq;
+
+	/* From in_flight test above, we know that cwnd > in_flight.  */
+	cong_win = (tp->snd_cwnd - in_flight) * tp->mss_cache;
+
+	limit = min(send_win, cong_win);
+
+	/* If sk_send_head can be sent fully now, just do it.  */
+	if (skb->len <= limit)
+		return 0;
+
+	if (sysctl_tcp_tso_win_divisor) {
+		u32 chunk = min(tp->snd_wnd, tp->snd_cwnd * tp->mss_cache);
+
+		/* If at least some fraction of a window is available,
+		 * just use it.
+		 */
+		chunk /= sysctl_tcp_tso_win_divisor;
+		if (limit >= chunk)
+			return 0;
+	} else {
+		/* Different approach, try not to defer past a single
+		 * ACK.  Receiver should ACK every other full sized
+		 * frame, so if we have space for more than 3 frames
+		 * then send now.
+		 */
+		if (limit > tcp_max_burst(tp) * tp->mss_cache)
+			return 0;
+	}
+
+	/* Ok, it looks like it is advisable to defer.  */
+	return 1;
+}
+
 /* This routine writes packets to the network.  It advances the
  * send_head.  This happens as incoming acks open up the remote
  * window for us.
@@ -887,8 +959,8 @@ static int tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct sk_buff *skb;
-	unsigned int tso_segs, cwnd_quota;
-	int sent_pkts;
+	unsigned int tso_segs, sent_pkts;
+	int cwnd_quota;
 
 	/* If we are closed, the bytes will have to remain here.
 	 * In time closedown will finish, we empty the write queue and all
@@ -903,24 +975,44 @@ static int tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle)
 
 	tso_segs = tcp_init_tso_segs(sk, skb);
 	cwnd_quota = tcp_cwnd_test(tp, skb);
+	if (unlikely(!cwnd_quota))
+		goto out;
+
 	sent_pkts = 0;
+	while (likely(tcp_snd_wnd_test(tp, skb, mss_now))) {
+		BUG_ON(!tso_segs);
 
-	while (cwnd_quota >= tso_segs) {
-		if (unlikely(!tcp_nagle_test(tp, skb, mss_now,
-					     (tcp_skb_is_last(sk, skb) ?
-					      nonagle : TCP_NAGLE_PUSH))))
-			break;
+		if (tso_segs == 1) {
+			if (unlikely(!tcp_nagle_test(tp, skb, mss_now,
+						     (tcp_skb_is_last(sk, skb) ?
+						      nonagle : TCP_NAGLE_PUSH))))
+				break;
+		} else {
+			if (tcp_tso_should_defer(sk, tp, skb))
+				break;
+		}
 
-		if (unlikely(!tcp_snd_wnd_test(tp, skb, mss_now)))
-			break;
+		if (tso_segs > 1) {
+			u32 limit = tcp_window_allows(tp, skb,
+						      mss_now, cwnd_quota);
+
+			if (skb->len < limit) {
+				unsigned int trim = skb->len % mss_now;
 
-		if (unlikely(skb->len > mss_now)) {
+				if (trim)
+					limit = skb->len - trim;
+			}
+			if (skb->len > limit) {
+				if (tso_fragment(sk, skb, limit))
+					break;
+			}
+		} else if (unlikely(skb->len > mss_now)) {
 			if (unlikely(tcp_fragment(sk, skb,  mss_now)))
 				break;
 		}
 
 		TCP_SKB_CB(skb)->when = tcp_time_stamp;
-		tcp_tso_set_push(skb);
+
 		if (unlikely(tcp_transmit_skb(sk, skb_clone(skb, GFP_ATOMIC))))
 			break;
 
@@ -936,6 +1028,11 @@ static int tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle)
 		 * the packet above, tso_segs will no longer be valid.
 		 */
 		cwnd_quota -= tcp_skb_pcount(skb);
+
+		BUG_ON(cwnd_quota < 0);
+		if (!cwnd_quota)
+			break;
+
 		skb = sk->sk_send_head;
 		if (!skb)
 			break;
@@ -946,7 +1043,7 @@ static int tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle)
 		tcp_cwnd_validate(sk, tp);
 		return 0;
 	}
-
+out:
 	return !tp->packets_out && sk->sk_send_head;
 }
 
@@ -965,6 +1062,53 @@ void __tcp_push_pending_frames(struct sock *sk, struct tcp_sock *tp,
 	}
 }
 
+/* Send _single_ skb sitting at the send head. This function requires
+ * true push pending frames to setup probe timer etc.
+ */
+void tcp_push_one(struct sock *sk, unsigned int mss_now)
+{
+	struct tcp_sock *tp = tcp_sk(sk);
+	struct sk_buff *skb = sk->sk_send_head;
+	unsigned int tso_segs, cwnd_quota;
+
+	BUG_ON(!skb || skb->len < mss_now);
+
+	tso_segs = tcp_init_tso_segs(sk, skb);
+	cwnd_quota = tcp_snd_test(sk, skb, mss_now, TCP_NAGLE_PUSH);
+
+	if (likely(cwnd_quota)) {
+		BUG_ON(!tso_segs);
+
+		if (tso_segs > 1) {
+			u32 limit = tcp_window_allows(tp, skb,
+						      mss_now, cwnd_quota);
+
+			if (skb->len < limit) {
+				unsigned int trim = skb->len % mss_now;
+
+				if (trim)
+					limit = skb->len - trim;
+			}
+			if (skb->len > limit) {
+				if (unlikely(tso_fragment(sk, skb, limit)))
+					return;
+			}
+		} else if (unlikely(skb->len > mss_now)) {
+			if (unlikely(tcp_fragment(sk, skb, mss_now)))
+				return;
+		}
+
+		/* Send it out now. */
+		TCP_SKB_CB(skb)->when = tcp_time_stamp;
+
+		if (likely(!tcp_transmit_skb(sk, skb_clone(skb, sk->sk_allocation)))) {
+			update_send_head(sk, tp, skb);
+			tcp_cwnd_validate(sk, tp);
+			return;
+		}
+	}
+}
+
 /* This function returns the amount that we can raise the
  * usable window based on the following constraints
  *  
@@ -1222,7 +1366,6 @@ int tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb)
 		if (sk->sk_route_caps & NETIF_F_TSO) {
 			sk->sk_route_caps &= ~NETIF_F_TSO;
 			sock_set_flag(sk, SOCK_NO_LARGESEND);
-			tp->mss_cache = tp->mss_cache_std;
 		}
 
 		if (tcp_trim_head(sk, skb, tp->snd_una - TCP_SKB_CB(skb)->seq))
@@ -1284,7 +1427,6 @@ int tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb)
 	 * is still in somebody's hands, else make a clone.
 	 */
 	TCP_SKB_CB(skb)->when = tcp_time_stamp;
-	tcp_tso_set_push(skb);
 
 	err = tcp_transmit_skb(sk, (skb_cloned(skb) ?
 				    pskb_copy(skb, GFP_ATOMIC):
@@ -1853,14 +1995,12 @@ int tcp_write_wakeup(struct sock *sk)
 				if (sk->sk_route_caps & NETIF_F_TSO) {
 					sock_set_flag(sk, SOCK_NO_LARGESEND);
 					sk->sk_route_caps &= ~NETIF_F_TSO;
-					tp->mss_cache = tp->mss_cache_std;
 				}
 			} else if (!tcp_skb_pcount(skb))
 				tcp_set_skb_tso_segs(sk, skb);
 
 			TCP_SKB_CB(skb)->flags |= TCPCB_FLAG_PSH;
 			TCP_SKB_CB(skb)->when = tcp_time_stamp;
-			tcp_tso_set_push(skb);
 			err = tcp_transmit_skb(sk, skb_clone(skb, GFP_ATOMIC));
 			if (!err) {
 				update_send_head(sk, tp, skb);
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 9dac7fdf4726..f6e288dc116e 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -2018,7 +2018,7 @@ static int tcp_v6_init_sock(struct sock *sk)
 	 */
 	tp->snd_ssthresh = 0x7fffffff;
 	tp->snd_cwnd_clamp = ~0;
-	tp->mss_cache_std = tp->mss_cache = 536;
+	tp->mss_cache = 536;
 
 	tp->reordering = sysctl_tcp_reordering;
 
-- 
cgit v1.2.3


From 6772926bef3c9f0ec761b39e5702535471fff70b Mon Sep 17 00:00:00 2001
From: Rusty Lynch <rusty.lynch@intel.com>
Date: Tue, 5 Jul 2005 18:54:50 -0700
Subject: [PATCH] kprobes: fix namespace problem and sparc64 build

The following renames arch_init, a kprobes function for performing any
architecture specific initialization, to arch_init_kprobes in order to
cleanup the namespace.

Also, this patch adds arch_init_kprobes to sparc64 to fix the sparc64 kprobes
build from the last return probe patch.

Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
---
 arch/i386/kernel/kprobes.c    | 2 +-
 arch/ia64/kernel/kprobes.c    | 2 +-
 arch/ppc64/kernel/kprobes.c   | 2 +-
 arch/sparc64/kernel/kprobes.c | 5 +++++
 arch/x86_64/kernel/kprobes.c  | 2 +-
 include/linux/kprobes.h       | 2 +-
 kernel/kprobes.c              | 2 +-
 7 files changed, 11 insertions(+), 6 deletions(-)

(limited to 'include/linux')

diff --git a/arch/i386/kernel/kprobes.c b/arch/i386/kernel/kprobes.c
index fc8b17521761..a6d8c45961d3 100644
--- a/arch/i386/kernel/kprobes.c
+++ b/arch/i386/kernel/kprobes.c
@@ -537,7 +537,7 @@ static struct kprobe trampoline_p = {
 	.pre_handler = trampoline_probe_handler
 };
 
-int __init arch_init(void)
+int __init arch_init_kprobes(void)
 {
 	return register_kprobe(&trampoline_p);
 }
diff --git a/arch/ia64/kernel/kprobes.c b/arch/ia64/kernel/kprobes.c
index 3aa3167edbec..884f5cd27d8a 100644
--- a/arch/ia64/kernel/kprobes.c
+++ b/arch/ia64/kernel/kprobes.c
@@ -713,7 +713,7 @@ static struct kprobe trampoline_p = {
 	.pre_handler = trampoline_probe_handler
 };
 
-int __init arch_init(void)
+int __init arch_init_kprobes(void)
 {
 	trampoline_p.addr =
 		(kprobe_opcode_t *)((struct fnptr *)kretprobe_trampoline)->ip;
diff --git a/arch/ppc64/kernel/kprobes.c b/arch/ppc64/kernel/kprobes.c
index 1d2ff6d6b0b3..a3d519518fb8 100644
--- a/arch/ppc64/kernel/kprobes.c
+++ b/arch/ppc64/kernel/kprobes.c
@@ -444,7 +444,7 @@ static struct kprobe trampoline_p = {
 	.pre_handler = trampoline_probe_handler
 };
 
-int __init arch_init(void)
+int __init arch_init_kprobes(void)
 {
 	return register_kprobe(&trampoline_p);
 }
diff --git a/arch/sparc64/kernel/kprobes.c b/arch/sparc64/kernel/kprobes.c
index bdac631cf011..bbf11f85dab1 100644
--- a/arch/sparc64/kernel/kprobes.c
+++ b/arch/sparc64/kernel/kprobes.c
@@ -433,3 +433,8 @@ int longjmp_break_handler(struct kprobe *p, struct pt_regs *regs)
 	return 0;
 }
 
+/* architecture specific initialization */
+int arch_init_kprobes(void)
+{
+	return 0;
+}
diff --git a/arch/x86_64/kernel/kprobes.c b/arch/x86_64/kernel/kprobes.c
index acd2a778ebe6..5c6dc7051482 100644
--- a/arch/x86_64/kernel/kprobes.c
+++ b/arch/x86_64/kernel/kprobes.c
@@ -682,7 +682,7 @@ static struct kprobe trampoline_p = {
 	.pre_handler = trampoline_probe_handler
 };
 
-int __init arch_init(void)
+int __init arch_init_kprobes(void)
 {
 	return register_kprobe(&trampoline_p);
 }
diff --git a/include/linux/kprobes.h b/include/linux/kprobes.h
index b7a194c4362a..e050fc2d4c26 100644
--- a/include/linux/kprobes.h
+++ b/include/linux/kprobes.h
@@ -155,7 +155,7 @@ extern void arch_copy_kprobe(struct kprobe *p);
 extern void arch_arm_kprobe(struct kprobe *p);
 extern void arch_disarm_kprobe(struct kprobe *p);
 extern void arch_remove_kprobe(struct kprobe *p);
-extern int arch_init(void);
+extern int arch_init_kprobes(void);
 extern void show_registers(struct pt_regs *regs);
 extern kprobe_opcode_t *get_insn_slot(void);
 extern void free_insn_slot(kprobe_opcode_t *slot);
diff --git a/kernel/kprobes.c b/kernel/kprobes.c
index 90c0e82b650c..b0237122b24e 100644
--- a/kernel/kprobes.c
+++ b/kernel/kprobes.c
@@ -574,7 +574,7 @@ static int __init init_kprobes(void)
 		INIT_HLIST_HEAD(&kretprobe_inst_table[i]);
 	}
 
-	err = arch_init();
+	err = arch_init_kprobes();
 	if (!err)
 		err = register_die_notifier(&kprobe_exceptions_nb);
 
-- 
cgit v1.2.3