Based on ideas of Andrew: http://marc.info/?l=linux-kernel&m=102912915020543&w=2 Scale the bdi dirty limit inversly with the tasks dirty rate. This makes heavy writers have a lower dirty limit than the occasional writer. Andrea proposed something similar: http://lwn.net/Articles/152277/ The main disadvantage to his patch is that he uses an unrelated quantity to measure time, which leaves him with a workload dependant tunable. Other than that the two approaches appear quite similar. Signed-off-by: Peter Zijlstra --- Changes since -v8: - initialized init_task - moved the prop_local init after the task_struct copy - changed the per task ratio to 1/8 (from 1/2). - explicit usage of prop_local_single include/linux/init_task.h | 1 include/linux/sched.h | 2 + kernel/fork.c | 10 +++++++++ mm/page-writeback.c | 50 +++++++++++++++++++++++++++++++++++++++++++++- 4 files changed, 62 insertions(+), 1 deletion(-) Index: linux-2.6/include/linux/sched.h =================================================================== --- linux-2.6.orig/include/linux/sched.h +++ linux-2.6/include/linux/sched.h @@ -84,6 +84,7 @@ struct sched_param { #include #include #include +#include #include @@ -1125,6 +1126,7 @@ struct task_struct { #ifdef CONFIG_FAULT_INJECTION int make_it_fail; #endif + struct prop_local_single dirties; }; /* Index: linux-2.6/kernel/fork.c =================================================================== --- linux-2.6.orig/kernel/fork.c +++ linux-2.6/kernel/fork.c @@ -107,6 +107,7 @@ static struct kmem_cache *mm_cachep; void free_task(struct task_struct *tsk) { + prop_local_destroy_single(&tsk->dirties); free_thread_info(tsk->stack); rt_mutex_debug_task_free(tsk); free_task_struct(tsk); @@ -163,6 +164,7 @@ static struct task_struct *dup_task_stru { struct task_struct *tsk; struct thread_info *ti; + int err; prepare_to_copy(orig); @@ -178,6 +180,14 @@ static struct task_struct *dup_task_stru *tsk = *orig; tsk->stack = ti; + + err = prop_local_init_single(&tsk->dirties); + if (err) { + free_thread_info(ti); + free_task_struct(tsk); + return NULL; + } + setup_thread_stack(tsk, orig); #ifdef CONFIG_CC_STACKPROTECTOR Index: linux-2.6/mm/page-writeback.c =================================================================== --- linux-2.6.orig/mm/page-writeback.c +++ linux-2.6/mm/page-writeback.c @@ -118,6 +118,7 @@ static void background_writeout(unsigned * */ static struct prop_descriptor vm_completions; +static struct prop_descriptor vm_dirties; static unsigned long determine_dirtyable_memory(void); @@ -146,6 +147,7 @@ int dirty_ratio_handler(ctl_table *table if (ret == 0 && write && vm_dirty_ratio != old_ratio) { int shift = calc_period_shift(); prop_change_shift(&vm_completions, shift); + prop_change_shift(&vm_dirties, shift); } return ret; } @@ -159,6 +161,11 @@ static inline void __bdi_writeout_inc(st __prop_inc_percpu(&vm_completions, &bdi->completions); } +static inline void task_dirty_inc(struct task_struct *tsk) +{ + prop_inc_single(&vm_dirties, &tsk->dirties); +} + /* * Obtain an accurate fraction of the BDI's portion. */ @@ -198,6 +205,37 @@ clip_bdi_dirty_limit(struct backing_dev_ *pbdi_dirty = min(*pbdi_dirty, avail_dirty); } +static inline void task_dirties_fraction(struct task_struct *tsk, + long *numerator, long *denominator) +{ + prop_fraction_single(&vm_dirties, &tsk->dirties, + numerator, denominator); +} + +/* + * scale the dirty limit + * + * task specific dirty limit: + * + * dirty -= (dirty/8) * p_{t} + */ +void task_dirty_limit(struct task_struct *tsk, long *pdirty) +{ + long numerator, denominator; + long dirty = *pdirty; + long long inv = dirty >> 3; + + task_dirties_fraction(tsk, &numerator, &denominator); + inv *= numerator; + do_div(inv, denominator); + + dirty -= inv; + if (dirty < *pdirty/2) + dirty = *pdirty/2; + + *pdirty = dirty; +} + /* * Work out the current dirty-memory clamping and background writeout * thresholds. @@ -304,6 +342,7 @@ get_dirty_limits(long *pbackground, long *pbdi_dirty = bdi_dirty; clip_bdi_dirty_limit(bdi, dirty, pbdi_dirty); + task_dirty_limit(current, pbdi_dirty); } } @@ -725,6 +764,7 @@ void __init page_writeback_init(void) shift = calc_period_shift(); prop_descriptor_init(&vm_completions, shift); + prop_descriptor_init(&vm_dirties, shift); } /** @@ -1003,7 +1043,7 @@ EXPORT_SYMBOL(redirty_page_for_writepage * If the mapping doesn't provide a set_page_dirty a_op, then * just fall through and assume that it wants buffer_heads. */ -int fastcall set_page_dirty(struct page *page) +static int __set_page_dirty(struct page *page) { struct address_space *mapping = page_mapping(page); @@ -1021,6 +1061,14 @@ int fastcall set_page_dirty(struct page } return 0; } + +int fastcall set_page_dirty(struct page *page) +{ + int ret = __set_page_dirty(page); + if (ret) + task_dirty_inc(current); + return ret; +} EXPORT_SYMBOL(set_page_dirty); /* Index: linux-2.6/include/linux/init_task.h =================================================================== --- linux-2.6.orig/include/linux/init_task.h +++ linux-2.6/include/linux/init_task.h @@ -169,6 +169,7 @@ extern struct group_info init_groups; [PIDTYPE_PGID] = INIT_PID_LINK(PIDTYPE_PGID), \ [PIDTYPE_SID] = INIT_PID_LINK(PIDTYPE_SID), \ }, \ + .dirties = INIT_PROP_LOCAL_SINGLE(dirties), \ INIT_TRACE_IRQFLAGS \ INIT_LOCKDEP \ } -- - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/