# This is a BitKeeper generated diff -Nru style patch. # # ChangeSet # 2004/12/13 08:49:27-05:00 jeffpc@optonline.net # cfq-iosched.c: # Fixup # # drivers/block/cfq-iosched.c # 2004/12/13 08:48:39-05:00 jeffpc@optonline.net +2 -0 # Fixup # # ChangeSet # 2004/12/12 23:54:22-05:00 jeffpc@optonline.net # cfq-iosched.c: # Changed slice lengths and scaling factors # # drivers/block/cfq-iosched.c # 2004/12/12 23:53:11-05:00 jeffpc@optonline.net +11 -11 # Changed slice lengths and scaling factors # # ChangeSet # 2004/12/12 22:17:02-05:00 jeffpc@optonline.net # Use iosched neutral defines # # init/main.c # 2004/12/12 22:16:50-05:00 jeffpc@optonline.net +1 -0 # Set current->io_prio to the default value # # include/linux/resource.h # 2004/12/12 22:16:50-05:00 jeffpc@optonline.net +4 -0 # IOscheduler neutral priority defines # # drivers/block/cfq-iosched.c # 2004/12/12 22:16:50-05:00 jeffpc@optonline.net +2 -2 # Use the iosched neutral defines # # ChangeSet # 2004/12/12 16:34:13-05:00 jeffpc@optonline.net # Fix # # kernel/sys.c # 2004/12/12 16:34:01-05:00 jeffpc@optonline.net +1 -1 # Fix # # kernel/sched.c # 2004/12/12 16:34:01-05:00 jeffpc@optonline.net +0 -9 # Remove ioprio from renicing-code # # ChangeSet # 2004/12/12 14:00:43-05:00 jeffpc@optonline.net # Clean up from a merge # # Signed-off-by: Josef "Jeff" Sipek # # drivers/block/cfq-iosched.c # 2004/12/12 14:00:30-05:00 jeffpc@optonline.net +0 -3 # Cleanup # # ChangeSet # 2004/12/12 13:54:13-05:00 jeffpc@optonline.net # Make io priorities 0..39 instead of a nice-like -20..19 # # Signed-off-by: Josef "Jeff" Sipek # # kernel/sys.c # 2004/12/12 13:54:00-05:00 jeffpc@optonline.net +7 -13 # Make io priorities 0..39 instead of a nice-like -20..19 # # kernel/sched.c # 2004/12/12 13:54:00-05:00 jeffpc@optonline.net +15 -16 # Make io priorities 0..39 instead of a nice-like -20..19 # # ChangeSet # 2004/12/12 11:52:09-05:00 jeffpc@optonline.net # Clean up unused variables # # Signed-off-by: Josef "Jeff" Sipek # # drivers/block/ll_rw_blk.c # 2004/12/12 11:51:57-05:00 jeffpc@optonline.net +0 -1 # Clean up unused variables # # drivers/block/cfq-iosched.c # 2004/12/12 11:51:57-05:00 jeffpc@optonline.net +0 -1 # Clean up unused variables # # ChangeSet # 2004/12/12 10:45:22-05:00 jeffpc@optonline.net # Manual/Auto merge # # drivers/block/cfq-iosched.c # 2004/12/12 10:45:13-05:00 jeffpc@optonline.net +0 -132 # Manual merge # # kernel/sched.c # 2004/12/12 10:34:17-05:00 jeffpc@optonline.net +0 -35 # Auto merged # # include/linux/sched.h # 2004/12/12 10:34:17-05:00 jeffpc@optonline.net +0 -1 # Auto merged # # include/linux/blkdev.h # 2004/12/12 10:34:17-05:00 jeffpc@optonline.net +0 -7 # Auto merged # # drivers/block/ll_rw_blk.c # 2004/12/12 10:34:17-05:00 jeffpc@optonline.net +0 -2 # Auto merged # # ChangeSet # 2004/12/11 12:42:25-05:00 jeffpc@optonline.net # Update to Jens Axboe's #4 version # # kernel/sched.c # 2004/12/11 11:55:49-05:00 jeffpc@optonline.net +35 -0 # Import patch n4-3 # # include/linux/sched.h # 2004/12/11 11:55:48-05:00 jeffpc@optonline.net +1 -0 # Import patch n4-3 # # include/linux/blkdev.h # 2004/12/11 11:55:49-05:00 jeffpc@optonline.net +7 -8 # Import patch n4-3 # # drivers/block/ll_rw_blk.c # 2004/12/11 11:55:49-05:00 jeffpc@optonline.net +3 -2 # Import patch n4-3 # # drivers/block/cfq-iosched.c # 2004/12/11 12:33:15-05:00 jeffpc@optonline.net +137 -74 # Import patch n4-3 # # ChangeSet # 2004/12/10 20:38:05-05:00 jeffpc@optonline.net # Added 2 syscalls sys_{get,set}iopriority(). Both 99% compied # from sys_{get,set}priority() # Added task_ioprio() # Little cleanup # # Signed-off-by: Josef "Jeff" Sipek # # kernel/sys.c # 2004/12/10 20:37:47-05:00 jeffpc@optonline.net +121 -10 # Little clean up # Added sys_{set,get}iopriority() # # kernel/sched.c # 2004/12/10 20:37:47-05:00 jeffpc@optonline.net +9 -1 # Tiny clean up # Added task_ioprio() # # include/linux/sched.h # 2004/12/10 20:37:47-05:00 jeffpc@optonline.net +1 -0 # Added prototype for task_ioprio() # # include/asm-i386/unistd.h # 2004/12/10 20:37:47-05:00 jeffpc@optonline.net +3 -1 # Added 2 syscalls sys_{get,set}iopriority(). Both 99% compied # from sys_{get,set}priority() # # arch/i386/kernel/entry.S # 2004/12/10 20:37:47-05:00 jeffpc@optonline.net +2 -0 # Added 2 syscalls sys_{get,set}iopriority(). Both 99% compied # from sys_{get,set}priority() # # ChangeSet # 2004/12/08 18:07:26-05:00 jeffpc@optonline.net # Changed sys_setpriority() # # Signed-off-by: Josef "Jeff" Sipek # # kernel/sys.c # 2004/12/08 18:07:11-05:00 jeffpc@optonline.net +26 -2 # Added set_one_ioprio() # Changed sys_setpriority() to modify io_prio as well # # kernel/sched.c # 2004/12/08 18:07:11-05:00 jeffpc@optonline.net +1 -1 # Display NICE-level ioprio # # include/linux/sched.h # 2004/12/08 18:07:11-05:00 jeffpc@optonline.net +1 -0 # Added set_user_ioprio() prototype # # ChangeSet # 2004/12/08 17:00:00-05:00 jeffpc@optonline.net # Added sys_io_prio() syscall # # Signed-off-by: Josef "Jeff" Sipek # # kernel/sched.c # 2004/12/08 16:59:46-05:00 jeffpc@optonline.net +37 -21 # Cleaned up set_user_ioprio() # Added sys_io_prio() # # include/asm-i386/unistd.h # 2004/12/08 16:59:45-05:00 jeffpc@optonline.net +2 -1 # Added sys_io_prio syscall number # # arch/i386/kernel/entry.S # 2004/12/08 16:59:45-05:00 jeffpc@optonline.net +1 -0 # Added sys_io_prio to sys_call_table # # ChangeSet # 2004/12/06 10:10:15-05:00 jeffpc@optonline.net # Fix compilation # # Signed-off-by: Josef "Jeff" Sipek # # kernel/sched.c # 2004/12/06 10:09:58-05:00 jeffpc@optonline.net +1 -3 # Fix compilation # # drivers/block/cfq-iosched.c # 2004/12/06 10:09:58-05:00 jeffpc@optonline.net +1 -0 # Fix compilation # # ChangeSet # 2004/12/05 22:11:23-05:00 jeffpc@optonline.net # Changed default sync and async cfq slice times # Added prio table to cfq # Added io_prio to task_struct # Added security_task_setioprio() # Added set_user_ioprio(), NICE_TO_IOPRIO(), IOPRIO_TO_NICE(), # TASK_IOPRIO(). Additionally, sys_nice() now changes io_prio # by the same increment as the nice value. # # Signed-off-by: Josef "Jeff" Sipek # # kernel/sched.c # 2004/12/05 22:11:13-05:00 jeffpc@optonline.net +40 -0 # Added set_user_ioprio(), NICE_TO_IOPRIO(), IOPRIO_TO_NICE(), # TASK_IOPRIO(). Additionally, sys_nice() now changes io_prio # by the same increment as the nice value. # # include/linux/security.h # 2004/12/05 22:11:13-05:00 jeffpc@optonline.net +12 -0 # Added security_task_setioprio # # include/linux/sched.h # 2004/12/05 22:11:13-05:00 jeffpc@optonline.net +3 -0 # Added io_prio to task_struct # # drivers/block/cfq-iosched.c # 2004/12/05 22:11:13-05:00 jeffpc@optonline.net +19 -3 # Changed the default sync and async slice time # Added prio/slice-scaling table # diff -Nru a/arch/i386/kernel/entry.S b/arch/i386/kernel/entry.S --- a/arch/i386/kernel/entry.S 2004-12-13 08:49:48 -05:00 +++ b/arch/i386/kernel/entry.S 2004-12-13 08:49:48 -05:00 @@ -864,5 +864,8 @@ .long sys_add_key .long sys_request_key .long sys_keyctl + .long sys_io_prio + .long sys_getiopriority /* 290 */ + .long sys_setiopriority syscall_table_size=(.-sys_call_table) diff -Nru a/drivers/block/cfq-iosched.c b/drivers/block/cfq-iosched.c --- a/drivers/block/cfq-iosched.c 2004-12-13 08:49:48 -05:00 +++ b/drivers/block/cfq-iosched.c 2004-12-13 08:49:48 -05:00 @@ -33,10 +33,20 @@ static int cfq_back_max = 16 * 1024; /* maximum backwards seek, in KiB */ static int cfq_back_penalty = 2; /* penalty of a backwards seek */ -static int cfq_slice_sync = HZ / 10; -static int cfq_slice_async = HZ / 25; +static int cfq_slice_sync = HZ / 1000; +static int cfq_slice_async = HZ / 1000; static int cfq_slice_async_rq = 16; -static int cfq_slice_idle = HZ / 249; +static int cfq_slice_idle = HZ / 1000; + +/* the slice scaling factors */ +static int cfq_prio_scale[] = {223, 194, 169, 147, 128, /* 0..4 */ + 111, 97, 84, 73, 64, /* 5..9 */ + 56, 49, 42, 36, 32, /* 10..14 */ + 28, 24, 21, 18, 16, /* 15..19 */ + 14, 12, 11, 9, 8, /* 20..24 */ + 7, 6, 5, 5, 4, /* 25..29 */ + 3, 3, 2, 2, 2, /* 30..34 */ + 1, 1, 1, 1, 1}; /* 35..39 */ static int cfq_max_depth = 4; @@ -98,6 +108,9 @@ */ #define CFQ_KEY_SPARE (~0UL) +#define CFQ_MIN_IO_PRIO IOPRIO_MIN +#define CFQ_MAX_IO_PRIO IOPRIO_MAX + static kmem_cache_t *crq_pool; static kmem_cache_t *cfq_pool; static kmem_cache_t *cfq_ioc_pool; @@ -133,6 +146,7 @@ struct work_struct unplug_work; struct cfq_queue *active_queue; unsigned int dispatch_slice; + unsigned int cfq_prio_scale[(CFQ_MAX_IO_PRIO - CFQ_MIN_IO_PRIO) + 1]; /* * tunables, see top of file @@ -176,6 +190,7 @@ unsigned long last_fifo_expire; int key_type; + int io_prio; unsigned long slice_start; unsigned long slice_end; @@ -912,7 +927,7 @@ * sync, use the sync time slice value */ if (!cfqq->slice_end) - cfqq->slice_end = cfqd->cfq_slice[!!sync] + jiffies; + cfqq->slice_end = cfqd->cfq_prio_scale[cfqq->io_prio]*cfqd->cfq_slice[!!sync] + jiffies; /* * expire an async queue immediately if it has used up its slice @@ -946,7 +961,6 @@ { struct cfq_queue *cfqq = crq->cfq_queue; struct cfq_data *cfqd = cfqq->cfqd; - unsigned long now, elapsed; if (unlikely(!blk_fs_request(crq->request))) return; @@ -972,13 +986,13 @@ WARN_ON(!cfqd->rq_in_driver); cfqd->rq_in_driver--; } - + /* * queue was preempted while this request was servicing */ if (cfqd->active_queue != cfqq) return; - + /* * this is still the active queue. if we have nothing to do and no * more pending requests in flight, wait for a new sync request if @@ -1093,6 +1107,7 @@ *cfqq = __cfqq; } + __cfqq->io_prio = current->io_prio; cic->cfqq = __cfqq; spin_unlock_irqrestore(cfqd->queue->queue_lock, flags); } @@ -1744,6 +1759,7 @@ cfqd->cfq_slice_async_rq = cfq_slice_async_rq; cfqd->cfq_slice_idle = cfq_slice_idle; cfqd->cfq_max_depth = cfq_max_depth; + memcpy(cfqd->cfq_prio_scale, cfq_prio_scale, sizeof(cfq_prio_scale)); return 0; out_spare: diff -Nru a/include/asm-i386/unistd.h b/include/asm-i386/unistd.h --- a/include/asm-i386/unistd.h 2004-12-13 08:49:48 -05:00 +++ b/include/asm-i386/unistd.h 2004-12-13 08:49:48 -05:00 @@ -294,8 +294,11 @@ #define __NR_add_key 286 #define __NR_request_key 287 #define __NR_keyctl 288 +#define __NR_io_prio 289 +#define __NR_getiopriority 290 +#define __NR_setiopriority 291 -#define NR_syscalls 289 +#define NR_syscalls 292 /* * user-visible error numbers are in the range -1 - -128: see diff -Nru a/include/linux/resource.h b/include/linux/resource.h --- a/include/linux/resource.h 2004-12-13 08:49:48 -05:00 +++ b/include/linux/resource.h 2004-12-13 08:49:48 -05:00 @@ -42,6 +42,10 @@ unsigned long rlim_max; }; +#define IOPRIO_MIN 0 +#define IOPRIO_MAX 39 +#define IOPRIO_DEFAULT 20 + #define PRIO_MIN (-20) #define PRIO_MAX 20 diff -Nru a/include/linux/sched.h b/include/linux/sched.h --- a/include/linux/sched.h 2004-12-13 08:49:48 -05:00 +++ b/include/linux/sched.h 2004-12-13 08:49:48 -05:00 @@ -664,6 +664,9 @@ struct mempolicy *mempolicy; short il_next; /* could be shared with used_math */ #endif + +/* io priority */ + short io_prio; }; static inline pid_t process_group(struct task_struct *tsk) @@ -723,8 +726,10 @@ extern void sched_idle_next(void); extern void set_user_nice(task_t *p, long nice); +extern void set_user_ioprio(task_t *p, long ioprio); extern int task_prio(const task_t *p); extern int task_nice(const task_t *p); +extern int task_ioprio(const task_t *p); extern int task_curr(const task_t *p); extern int idle_cpu(int cpu); extern unsigned long task_will_schedule_at(const task_t *p); diff -Nru a/include/linux/security.h b/include/linux/security.h --- a/include/linux/security.h 2004-12-13 08:49:48 -05:00 +++ b/include/linux/security.h 2004-12-13 08:49:48 -05:00 @@ -1717,6 +1717,13 @@ return security_ops->task_setnice (p, nice); } +static inline int security_task_setioprio (struct task_struct *p, int prio) +{ + #warn "Fixme! Add security_ops->task_setprio" + return 1; + //return security_ops->task_setioprio (p, prio); +} + static inline int security_task_setrlimit (unsigned int resource, struct rlimit *new_rlim) { @@ -2353,6 +2360,11 @@ } static inline int security_task_setnice (struct task_struct *p, int nice) +{ + return 0; +} + +static inline int security_task_setioprio (struct task_struct *p, int prio) { return 0; } diff -Nru a/init/main.c b/init/main.c --- a/init/main.c 2004-12-13 08:49:48 -05:00 +++ b/init/main.c 2004-12-13 08:49:48 -05:00 @@ -443,6 +443,7 @@ static void noinline rest_init(void) __releases(kernel_lock) { + current->io_prio = IOPRIO_DEFAULT; kernel_thread(init, NULL, CLONE_FS | CLONE_SIGHAND); numa_default_policy(); unlock_kernel(); diff -Nru a/kernel/sched.c b/kernel/sched.c --- a/kernel/sched.c 2004-12-13 08:49:48 -05:00 +++ b/kernel/sched.c 2004-12-13 08:49:48 -05:00 @@ -16,6 +16,7 @@ * by Davide Libenzi, preemptible kernel bits by Robert Love. * 2003-09-03 Interactivity tuning by Con Kolivas. * 2004-04-02 Scheduler domains code by Nick Piggin + * 2004-12-07 IO Priorities */ #include @@ -63,6 +64,8 @@ #define PRIO_TO_NICE(prio) ((prio) - MAX_RT_PRIO - 20) #define TASK_NICE(p) PRIO_TO_NICE((p)->static_prio) +#define TASK_IOPRIO(p) ((p)->io_prio) + /* * 'User priority' is the nice value converted to something we * can work with better when scaling various scheduler parameters, @@ -3024,6 +3027,17 @@ EXPORT_SYMBOL(set_user_nice); +void set_user_ioprio(task_t *p, long ioprio) +{ + if (TASK_IOPRIO(p) == ioprio || ioprio < 0 || ioprio > 39) + return; + + p->io_prio = ioprio; + printk(KERN_ERR "task %d now has io_prio = %d\n", p->pid, p->io_prio); +} + +EXPORT_SYMBOL(set_user_ioprio); + #ifdef __ARCH_WANT_SYS_NICE /* @@ -3068,6 +3082,35 @@ #endif +/* + * sys_ioprio - change the io priority of the current process. + * @increment: priority increment + */ +asmlinkage long sys_io_prio(int increment) +{ + long io_prio; + + if (increment < 0) { + if (!capable(CAP_SYS_NICE)) + return -EPERM; + if (increment < -40) + increment = -40; + } + if (increment > 40) + increment = 40; + + io_prio = current->io_prio + increment; + if (io_prio < 0) + io_prio = 0; + if (io_prio > 39) + io_prio = 39; + +#warning "Fixme: add call to security_task_setioprio()" + + set_user_ioprio(current, io_prio); + return 0; +} + /** * task_prio - return the priority value of a given task. * @p: the task in question. @@ -3088,6 +3131,15 @@ int task_nice(const task_t *p) { return TASK_NICE(p); +} + +/** + * task_ioprio - return the ioprio value of a given task. + * @p: the task in question. + */ +int task_ioprio(const task_t *p) +{ + return TASK_IOPRIO(p); } /** diff -Nru a/kernel/sys.c b/kernel/sys.c --- a/kernel/sys.c 2004-12-13 08:49:48 -05:00 +++ b/kernel/sys.c 2004-12-13 08:49:48 -05:00 @@ -240,6 +240,23 @@ return error; } +static int set_one_ioprio(struct task_struct *p, int prioval, int error) +{ + if (p->uid != current->euid && + p->uid != current->uid && !capable(CAP_SYS_NICE)) { + error = -EPERM; + goto out; + } + if (prioval < task_ioprio(p) && !capable(CAP_SYS_NICE)) { + error = -EACCES; + goto out; + } + #warning "FIXME: call security_task_setioprio(p, prioval);" + set_user_ioprio(p, prioval); +out: + return error; +} + asmlinkage long sys_setpriority(int which, int who, int niceval) { struct task_struct *g, *p; @@ -357,6 +374,118 @@ return retval; } +asmlinkage long sys_setiopriority(int which, int who, int prioval) +{ + struct task_struct *g, *p; + struct user_struct *user; + int error = -EINVAL; + + if (which > 2 || which < 0) + goto out; + + /* normalize: avoid signed division (rounding problems) */ + error = -ESRCH; + if (prioval < 0) + prioval = 0; + if (prioval > 39) + prioval = 39; + + read_lock(&tasklist_lock); + switch (which) { + case PRIO_PROCESS: + if (!who) + who = current->pid; + p = find_task_by_pid(who); + if (p) { + error = set_one_ioprio(p, prioval, error); + } + break; + case PRIO_PGRP: + if (!who) + who = process_group(current); + do_each_task_pid(who, PIDTYPE_PGID, p) { + error = set_one_ioprio(p, prioval, error); + } while_each_task_pid(who, PIDTYPE_PGID, p); + break; + case PRIO_USER: + if (!who) + user = current->user; + else + user = find_user(who); + + if (!user) + goto out_unlock; + + do_each_thread(g, p) + if (p->uid == who) { + error = set_one_ioprio(p, prioval, error); + } + while_each_thread(g, p); + if (who) + free_uid(user); /* For find_user() */ + break; + } +out_unlock: + read_unlock(&tasklist_lock); +out: + return error; +} + +asmlinkage long sys_getiopriority(int which, int who) +{ + struct task_struct *g, *p; + struct user_struct *user; + long prioval, retval = -ESRCH; + + if (which > 2 || which < 0) + return -EINVAL; + + read_lock(&tasklist_lock); + switch (which) { + case PRIO_PROCESS: + if (!who) + who = current->pid; + p = find_task_by_pid(who); + if (p) { + prioval = task_ioprio(p); + if (prioval > retval) + retval = prioval; + } + break; + case PRIO_PGRP: + if (!who) + who = process_group(current); + do_each_task_pid(who, PIDTYPE_PGID, p) { + prioval = task_ioprio(p); + if (prioval > retval) + retval = prioval; + } while_each_task_pid(who, PIDTYPE_PGID, p); + break; + case PRIO_USER: + if (!who) + user = current->user; + else + user = find_user(who); + + if (!user) + goto out_unlock; + + do_each_thread(g, p) + if (p->uid == who) { + prioval = task_ioprio(p); + if (prioval > retval) + retval = prioval; + } + while_each_thread(g, p); + if (who) + free_uid(user); /* for find_user() */ + break; + } +out_unlock: + read_unlock(&tasklist_lock); + + return retval; +} /* * Reboot system call: for obvious reasons only root may call it,