qsnet-rhel4-2.6.patch

lustre 1.6.5 source code

+#include <linux/ptrack.h>
 
 #define INIT_FILES \
 { 							\
@@ -112,6 +113,7 @@ extern struct group_info init_groups;
 	.proc_lock	= SPIN_LOCK_UNLOCKED,				\
 	.switch_lock	= SPIN_LOCK_UNLOCKED,				\
 	.journal_info	= NULL,						\
+	INIT_TASK_PTRACK(tsk)						\
 }
 
 
Index: linux-269-5502/include/linux/ioproc.h
===================================================================
--- /dev/null
+++ linux-269-5502/include/linux/ioproc.h
@@ -0,0 +1,270 @@
+/*
+ *    Copyright (C) 2006 Quadrics Ltd
+ *
+ *    This program is free software; you can redistribute it and/or modify
+ *    it under the terms of the GNU General Public License as published by
+ *    the Free Software Foundation; either version 2 of the License, or
+ *    (at your option) any later version.
+ *
+ *    This program is distributed in the hope that it will be useful,
+ *    but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *    GNU General Public License for more details.
+ *
+ *    You should have received a copy of the GNU General Public License
+ *    along with this program; if not, write to the Free Software
+ *    Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+/*
+ * Callbacks for IO processor page table updates.
+ */
+
+#ifndef __LINUX_IOPROC_H__
+#define __LINUX_IOPROC_H__
+
+#include <linux/sched.h>
+#include <linux/mm.h>
+
+typedef struct ioproc_ops {
+	struct ioproc_ops *next;
+	void *arg;
+
+	void (*release) (void *arg, struct mm_struct * mm);
+	void (*sync_range) (void *arg, struct vm_area_struct * vma,
+			    unsigned long start, unsigned long end);
+	void (*invalidate_range) (void *arg, struct vm_area_struct * vma,
+				  unsigned long start, unsigned long end);
+	void (*update_range) (void *arg, struct vm_area_struct * vma,
+			      unsigned long start, unsigned long end);
+
+	void (*change_protection) (void *arg, struct vm_area_struct * vma,
+				   unsigned long start, unsigned long end,
+				   pgprot_t newprot);
+
+	void (*sync_page) (void *arg, struct vm_area_struct * vma,
+			   unsigned long address);
+	void (*invalidate_page) (void *arg, struct vm_area_struct * vma,
+				 unsigned long address);
+	void (*update_page) (void *arg, struct vm_area_struct * vma,
+			     unsigned long address);
+
+} ioproc_ops_t;
+
+/* IOPROC Registration
+ * 
+ * Called by the IOPROC device driver to register its interest in page table
+ * changes for the process associated with the supplied mm_struct
+ *
+ * The caller should first allocate and fill out an ioproc_ops structure with 
+ * the function pointers initialised to the device driver specific code for
+ * each callback. If the device driver doesn't have code for a particular 
+ * callback then it should set the function pointer to be NULL.
+ * The ioproc_ops arg parameter will be passed unchanged as the first argument
+ * to each callback function invocation.
+ *
+ * The ioproc registration is not inherited across fork() and should be called
+ * once for each process that the IOPROC device driver is interested in.
+ *
+ * Must be called holding the mm->page_table_lock
+ */
+extern int ioproc_register_ops(struct mm_struct *mm, struct ioproc_ops *ip);
+
+/* IOPROC De-registration
+ * 
+ * Called by the IOPROC device driver when it is no longer interested in page 
+ * table changes for the process associated with the supplied mm_struct
+ *
+ * Normally this is not needed to be called as the ioproc_release() code will
+ * automatically unlink the ioproc_ops struct from the mm_struct as the
+ * process exits
+ *
+ * Must be called holding the mm->page_table_lock
+ */
+extern int ioproc_unregister_ops(struct mm_struct *mm, struct ioproc_ops *ip);
+
+#ifdef CONFIG_IOPROC
+
+/* IOPROC Release
+ *
+ * Called during exit_mmap() as all vmas are torn down and unmapped.
+ *
+ * Also unlinks the ioproc_ops structure from the mm list as it goes.
+ *
+ * No need for locks as the mm can no longer be accessed at this point
+ *
+ */
+static inline void ioproc_release(struct mm_struct *mm)
+{
+	struct ioproc_ops *cp;
+
+	while ((cp = mm->ioproc_ops) != NULL) {
+		mm->ioproc_ops = cp->next;
+
+		if (cp->release)
+			cp->release(cp->arg, mm);
+	}
+}
+
+/* IOPROC SYNC RANGE
+ *
+ * Called when a memory map is synchronised with its disk image i.e. when the 
+ * msync() syscall is invoked. Any future read or write to the associated 
+ * pages by the IOPROC should cause the page to be marked as referenced or 
+ * modified.
+ *
+ * Called holding the mm->page_table_lock
+ */
+static inline void
+ioproc_sync_range(struct vm_area_struct *vma, unsigned long start,
+		  unsigned long end)
+{
+	struct ioproc_ops *cp;
+
+	for (cp = vma->vm_mm->ioproc_ops; cp; cp = cp->next)
+		if (cp->sync_range)
+			cp->sync_range(cp->arg, vma, start, end);
+}
+
+/* IOPROC INVALIDATE RANGE
+ *
+ * Called whenever a valid PTE is unloaded e.g. when a page is unmapped by the
+ * user or paged out by the kernel. 
+ *
+ * After this call the IOPROC must not access the physical memory again unless
+ * a new translation is loaded.
+ *
+ * Called holding the mm->page_table_lock
+ */
+static inline void
+ioproc_invalidate_range(struct vm_area_struct *vma, unsigned long start,
+			unsigned long end)
+{
+	struct ioproc_ops *cp;
+
+	for (cp = vma->vm_mm->ioproc_ops; cp; cp = cp->next)
+		if (cp->invalidate_range)
+			cp->invalidate_range(cp->arg, vma, start, end);
+}
+
+/* IOPROC UPDATE RANGE
+ *
+ * Called whenever a valid PTE is loaded e.g. mmaping memory, moving the brk 
+ * up, when breaking COW or faulting in an anonymous page of memory.
+ *
+ * These give the IOPROC device driver the opportunity to load translations 
+ * speculatively, which can improve performance by avoiding device translation
+ * faults.
+ *
+ * Called holding the mm->page_table_lock
+ */
+static inline void
+ioproc_update_range(struct vm_area_struct *vma, unsigned long start,
+		    unsigned long end)
+{
+	struct ioproc_ops *cp;
+
+	for (cp = vma->vm_mm->ioproc_ops; cp; cp = cp->next)
+		if (cp->update_range)
+			cp->update_range(cp->arg, vma, start, end);
+}
+
+/* IOPROC CHANGE PROTECTION
+ *
+ * Called when the protection on a region of memory is changed i.e. when the 
+ * mprotect() syscall is invoked.
+ *
+ * The IOPROC must not be able to write to a read-only page, so if the 
+ * permissions are downgraded then it must honour them. If they are upgraded 
+ * it can treat this in the same way as the ioproc_update_[range|sync]() calls
+ *
+ * Called holding the mm->page_table_lock
+ */
+static inline void
+ioproc_change_protection(struct vm_area_struct *vma, unsigned long start,
+			 unsigned long end, pgprot_t newprot)
+{
+	struct ioproc_ops *cp;
+
+	for (cp = vma->vm_mm->ioproc_ops; cp; cp = cp->next)
+		if (cp->change_protection)
+			cp->change_protection(cp->arg, vma, start, end,
+					      newprot);
+}
+
+/* IOPROC SYNC PAGE
+ *
+ * Called when a memory map is synchronised with its disk image i.e. when the 
+ * msync() syscall is invoked. Any future read or write to the associated page
+ * by the IOPROC should cause the page to be marked as referenced or modified.
+ *
+ * Not currently called as msync() calls ioproc_sync_range() instead
+ *
+ * Called holding the mm->page_table_lock
+ */
+static inline void
+ioproc_sync_page(struct vm_area_struct *vma, unsigned long addr)
+{
+	struct ioproc_ops *cp;
+
+	for (cp = vma->vm_mm->ioproc_ops; cp; cp = cp->next)
+		if (cp->sync_page)
+			cp->sync_page(cp->arg, vma, addr);
+}
+
+/* IOPROC INVALIDATE PAGE
+ *
+ * Called whenever a valid PTE is unloaded e.g. when a page is unmapped by the
+ * user or paged out by the kernel. 
+ *
+ * After this call the IOPROC must not access the physical memory again unless
+ * a new translation is loaded.
+ *
+ * Called holding the mm->page_table_lock
+ */
+static inline void
+ioproc_invalidate_page(struct vm_area_struct *vma, unsigned long addr)
+{
+	struct ioproc_ops *cp;
+
+	for (cp = vma->vm_mm->ioproc_ops; cp; cp = cp->next)
+		if (cp->invalidate_page)
+			cp->invalidate_page(cp->arg, vma, addr);
+}
+
+/* IOPROC UPDATE PAGE
+ *
+ * Called whenever a valid PTE is loaded e.g. mmaping memory, moving the brk 
+ * up, when breaking COW or faulting in an anoymous page of memory.
+ *
+ * These give the IOPROC device the opportunity to load translations 
+ * speculatively, which can improve performance by avoiding device translation
+ * faults.
+ *
+ * Called holding the mm->page_table_lock
+ */
+static inline void
+ioproc_update_page(struct vm_area_struct *vma, unsigned long addr)
+{
+	struct ioproc_ops *cp;
+
+	for (cp = vma->vm_mm->ioproc_ops; cp; cp = cp->next)
+		if (cp->update_page)
+			cp->update_page(cp->arg, vma, addr);
+}
+
+#else
+
+/* ! CONFIG_IOPROC so make all hooks empty */
+
+#define ioproc_release(mm)				do { } while (0)
+#define ioproc_sync_range(vma, start, end)		do { } while (0)
+#define ioproc_invalidate_range(vma, start, end)	do { } while (0)
+#define ioproc_update_range(vma, start, end)		do { } while (0)
+#define ioproc_change_protection(vma, start, end, prot)	do { } while (0)
+#define ioproc_sync_page(vma, addr)			do { } while (0)
+#define ioproc_invalidate_page(vma, addr)		do { } while (0)
+#define ioproc_update_page(vma, addr)			do { } while (0)
+
+#endif				/* CONFIG_IOPROC */
+#endif				/* __LINUX_IOPROC_H__ */
Index: linux-269-5502/include/linux/sched.h
===================================================================
--- linux-269-5502.orig/include/linux/sched.h
+++ linux-269-5502/include/linux/sched.h
@@ -185,6 +185,9 @@ extern signed long schedule_timeout_unin
 asmlinkage void schedule(void);
 
 struct namespace;
+#ifdef CONFIG_IOPROC
+struct ioproc_ops;
+#endif
 
 /* Maximum number of active map areas.. This is a random (large) number */
 #define DEFAULT_MAX_MAP_COUNT	65536
@@ -260,6 +263,11 @@ struct mm_struct {
 	struct kioctx		*ioctx_list;
 
 	struct kioctx		default_kioctx;
+ 
+#ifdef CONFIG_IOPROC
+ 	/* hooks for io devices with advanced RDMA capabilities */
+	struct ioproc_ops       *ioproc_ops;
+#endif
 };
 
 extern int mmlist_nr;
@@ -635,6 +643,10 @@ struct task_struct {
   	struct mempolicy *mempolicy;
   	short il_next;		/* could be shared with used_math */
 #endif
+#ifdef CONFIG_PTRACK
+/* process tracking callback */
+	struct list_head ptrack_list;
+#endif
 };
 
 static inline pid_t process_group(struct task_struct *tsk)
Index: linux-269-5502/include/linux/ptrack.h
===================================================================
--- /dev/null
+++ linux-269-5502/include/linux/ptrack.h
@@ -0,0 +1,65 @@
+/*
+ *    Copyright (C) 2000  Regents of the University of California
+ *
+ *    This program is free software; you can redistribute it and/or modify
+ *    it under the terms of the GNU General Public License as published by
+ *    the Free Software Foundation; either version 2 of the License, or
+ *    (at your option) any later version.
+ *
+ *    This program is distributed in the hope that it will be useful,
+ *    but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *    GNU General Public License for more details.
+ *
+ *    You should have received a copy of the GNU General Public License
+ *    along with this program; if not, write to the Free Software
+ *    Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ *
+ * Derived from exit_actn.c by
+ *    Copyright (C) 2003 Quadrics Ltd.
+ *
+ */
+#ifndef __LINUX_PTRACK_H
+#define __LINUX_PTRACK_H
+
+/* 
+ * Process tracking - this allows a module to keep track of processes
+ * in order that it can manage all tasks derived from a single process.
+ */
+
+#define PTRACK_PHASE_CLONE	1
+#define PTRACK_PHASE_CLONE_FAIL	2
+#define PTRACK_PHASE_EXEC	3
+#define PTRACK_PHASE_EXIT      	4
+
+#define PTRACK_FINISHED		0
+#define PTRACK_INNHERIT		1
+#define PTRACK_DENIED		2
+
+#ifdef CONFIG_PTRACK
+
+typedef int (*ptrack_callback_t)(void *arg, int phase, struct task_struct *child);
+
+struct ptrack_desc {
+       struct list_head        link;
+       ptrack_callback_t       callback;
+       void                   *arg;
+};
+
+extern int     ptrack_register (ptrack_callback_t callback, void *arg);
+extern void    ptrack_deregister (ptrack_callback_t callback, void *arg);
+extern int     ptrack_registered (ptrack_callback_t callback, void *arg);
+
+extern int     ptrack_call_callbacks (int phase, struct task_struct *child);
+
+#define INIT_TASK_PTRACK(tsk) \
+	.ptrack_list = LIST_HEAD_INIT(tsk.ptrack_list)
+
+#else
+#define ptrack_call_callbacks (phase, child) (0)
+
+#define INIT_TASK_PTRACK(tsk)
+
+#endif
+
+#endif /* __LINUX_PTRACK_H */
Index: linux-269-5502/include/asm-ia64/param.h
===================================================================
--- linux-269-5502.orig/include/asm-ia64/param.h
+++ linux-269-5502/include/asm-ia64/param.h
@@ -27,7 +27,7 @@
    */
 #  define HZ	  32
 # else
-#  define HZ	1024
+#  define HZ	100
 # endif
 # define USER_HZ	HZ
 # define CLOCKS_PER_SEC	HZ	/* frequency at which times() counts */
Index: linux-269-5502/include/asm-i386/param.h
===================================================================
--- linux-269-5502.orig/include/asm-i386/param.h
+++ linux-269-5502/include/asm-i386/param.h
@@ -2,7 +2,7 @@
 #define _ASMi386_PARAM_H
 
 #ifdef __KERNEL__
-# define HZ		1000		/* Internal kernel timer frequency */
+# define HZ		100		/* Internal kernel timer frequency */
 # define USER_HZ	100		/* .. some user interfaces are in "ticks" */
 # define CLOCKS_PER_SEC		(USER_HZ)	/* like times() */
 #endif
Index: linux-269-5502/include/asm-x86_64/param.h
===================================================================
--- linux-269-5502.orig/include/asm-x86_64/param.h
+++ linux-269-5502/include/asm-x86_64/param.h
@@ -2,7 +2,7 @@
 #define _ASMx86_64_PARAM_H
 
 #ifdef __KERNEL__
-# define HZ            1000            /* Internal kernel timer frequency */
+# define HZ            100            /* Internal kernel timer frequency */
 # define USER_HZ       100          /* .. some user interfaces are in "ticks */
 #define CLOCKS_PER_SEC        (USER_HZ)       /* like times() */
 #endif