The patch titled mm: dirty balancing for tasks has been added to the -mm tree. Its filename is mm-dirty-balancing-for-tasks.patch *** Remember to use Documentation/SubmitChecklist when testing your code *** See http://www.zip.com.au/~akpm/linux/patches/stuff/added-to-mm.txt to find out what to do about this ------------------------------------------------------ Subject: mm: dirty balancing for tasks From: Peter Zijlstra <a.p.zijlstra@xxxxxxxxx> 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 <a.p.zijlstra@xxxxxxxxx> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> --- 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(-) diff -puN include/linux/init_task.h~mm-dirty-balancing-for-tasks include/linux/init_task.h --- a/include/linux/init_task.h~mm-dirty-balancing-for-tasks +++ a/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 \ } diff -puN include/linux/sched.h~mm-dirty-balancing-for-tasks include/linux/sched.h --- a/include/linux/sched.h~mm-dirty-balancing-for-tasks +++ a/include/linux/sched.h @@ -84,6 +84,7 @@ struct sched_param { #include <linux/timer.h> #include <linux/hrtimer.h> #include <linux/task_io_accounting.h> +#include <linux/proportions.h> #include <asm/processor.h> @@ -1123,6 +1124,7 @@ struct task_struct { #ifdef CONFIG_FAULT_INJECTION int make_it_fail; #endif + struct prop_local_single dirties; }; /* diff -puN kernel/fork.c~mm-dirty-balancing-for-tasks kernel/fork.c --- a/kernel/fork.c~mm-dirty-balancing-for-tasks +++ a/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 diff -puN mm/page-writeback.c~mm-dirty-balancing-for-tasks mm/page-writeback.c --- a/mm/page-writeback.c~mm-dirty-balancing-for-tasks +++ a/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(struct ctl_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. @@ -303,6 +341,7 @@ get_dirty_limits(long *pbackground, long *pbdi_dirty = bdi_dirty; clip_bdi_dirty_limit(bdi, dirty, pbdi_dirty); + task_dirty_limit(current, pbdi_dirty); } } @@ -724,6 +763,7 @@ void __init page_writeback_init(void) shift = calc_period_shift(); prop_descriptor_init(&vm_completions, shift); + prop_descriptor_init(&vm_dirties, shift); } /** @@ -1002,7 +1042,7 @@ 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); @@ -1020,6 +1060,14 @@ 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); /* _ Patches currently in -mm which might be from a.p.zijlstra@xxxxxxxxx are radix-tree-use-indirect-bit.patch nfs-remove-congestion_end.patch lib-percpu_counter_add.patch lib-percpu_counter_sub.patch lib-percpu_counter-variable-batch.patch lib-make-percpu_counter_add-take-s64.patch lib-percpu_counter_set.patch lib-percpu_counter_sum_positive.patch lib-percpu_count_sum.patch lib-percpu_counter_init-error-handling.patch lib-percpu_counter_init_irq.patch mm-bdi-init-hooks.patch mtd-bdi-init-hooks.patch mtd-clean-up-the-backing_dev_info-usage.patch mtd-give-mtdconcat-devices-their-own-backing_dev_info.patch mm-scalable-bdi-statistics-counters.patch mm-count-reclaimable-pages-per-bdi.patch mm-count-writeback-pages-per-bdi.patch mm-expose-bdi-statistics-in-sysfs.patch lib-floating-proportions.patch mm-per-device-dirty-threshold.patch mm-dirty-balancing-for-tasks.patch debug-sysfs-files-for-the-current-ratio-size-total.patch intel-iommu-dmar-detection-and-parsing-logic.patch intel-iommu-pci-generic-helper-function.patch intel-iommu-clflush_cache_range-now-takes-size-param.patch intel-iommu-iova-allocation-and-management-routines.patch intel-iommu-intel-iommu-driver.patch intel-iommu-avoid-memory-allocation-failures-in-dma-map-api-calls.patch intel-iommu-intel-iommu-cmdline-option-forcedac.patch intel-iommu-dmar-fault-handling-support.patch intel-iommu-iommu-gfx-workaround.patch intel-iommu-iommu-floppy-workaround.patch task-containersv11-shared-container-subsystem-group-arrays-avoid-lockdep-warning.patch task-containersv11-shared-container-subsystem-group-arrays-include-fix.patch containers-bdi-init-hooks.patch workqueue-debug-flushing-deadlocks-with-lockdep.patch workqueue-debug-work-related-deadlocks-with-lockdep.patch memory-controller-add-documentation.patch memory-controller-resource-counters-v7.patch memory-controller-containers-setup-v7.patch memory-controller-accounting-setup-v7.patch memory-controller-memory-accounting-v7.patch memory-controller-task-migration-v7.patch memory-controller-add-per-container-lru-and-reclaim-v7.patch memory-controller-add-per-container-lru-and-reclaim-v7-fix.patch memory-controller-improve-user-interface.patch memory-controller-oom-handling-v7.patch memory-controller-add-switch-to-control-what-type-of-pages-to-limit-v7.patch memory-controller-make-page_referenced-container-aware-v7.patch memory-controller-make-charging-gfp-mask-aware.patch - To unsubscribe from this list: send the line "unsubscribe mm-commits" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html