damon_call() works asynchronously and synchronously for repeat and non-repeat mode requests, respectively. The comment about the behavior is wrong, though. Fix it. The wrong comment was introduced together with the repeat mode, by commit 43df7676e550 ("mm/damon/core: introduce repeat mode damon_call()"). Signed-off-by: SeongJae Park --- mm/damon/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/damon/core.c b/mm/damon/core.c index d78f4452e536..06ad359024ad 100644 --- a/mm/damon/core.c +++ b/mm/damon/core.c @@ -1448,7 +1448,7 @@ bool damon_is_running(struct damon_ctx *ctx) * Ask DAMON worker thread (kdamond) of @ctx to call a function with an * argument data that respectively passed via &damon_call_control->fn and * &damon_call_control->data of @control. If &damon_call_control->repeat of - * @control is set, further wait until the kdamond finishes handling of the + * @control is unset, further wait until the kdamond finishes handling of the * request. Otherwise, return as soon as the request is made. * * The kdamond executes the function with the argument in the main loop, just -- 2.47.3 Commit b243d666d107 ("Docs/admin-guide/mm/damon/usage: add intervals_goal directory on the hierarchy") mistakenly added a wrong reference for intervals goal usage documentation on the design document. Fix it. Signed-off-by: SeongJae Park --- Documentation/mm/damon/design.rst | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/Documentation/mm/damon/design.rst b/Documentation/mm/damon/design.rst index b54925ea78e9..2d8d8ca1e0a3 100644 --- a/Documentation/mm/damon/design.rst +++ b/Documentation/mm/damon/design.rst @@ -381,8 +381,8 @@ That is, assumes 4% (20% of 20%) DAMON-observed access events ratio (source) to capture 64% (80% multipled by 80%) real access events (outcomes). To know how user-space can use this feature via :ref:`DAMON sysfs interface -`, refer to :ref:`intervals_goal ` part of -the documentation. +`, refer to :ref:`intervals_goal +` part of the documentation. .. _damon_design_damos: -- 2.47.3 It is a contextual typo. Fix it. Signed-off-by: SeongJae Park --- Documentation/admin-guide/mm/damon/stat.rst | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Documentation/admin-guide/mm/damon/stat.rst b/Documentation/admin-guide/mm/damon/stat.rst index 4c517c2c219a..20f540a9d3d2 100644 --- a/Documentation/admin-guide/mm/damon/stat.rst +++ b/Documentation/admin-guide/mm/damon/stat.rst @@ -17,7 +17,7 @@ DAMON_STAT uses monitoring intervals :ref:`auto-tuning ` to make its accuracy high and overhead minimum. It auto-tunes the intervals aiming 4 % of observable access events to be captured in each snapshot, while limiting the resulting sampling -events to be 5 milliseconds in minimum and 10 seconds in maximum. On a few +interval to be 5 milliseconds in minimum and 10 seconds in maximum. On a few production server systems, it resulted in consuming only 0.x % single CPU time, while capturing reasonable quality of access patterns. -- 2.47.3 Committing a monitoring target with empty target regions is for keeping the current monitoring results. This behavior was introduced by commit 973233600676 ("mm/damon/sysfs: update monitoring target regions for online input commit"). The behavior is not documented, though. Update the usage document for clarifying this behavior. Signed-off-by: SeongJae Park --- Documentation/admin-guide/mm/damon/usage.rst | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/Documentation/admin-guide/mm/damon/usage.rst b/Documentation/admin-guide/mm/damon/usage.rst index d8a3d6e740b3..9991dad60fcf 100644 --- a/Documentation/admin-guide/mm/damon/usage.rst +++ b/Documentation/admin-guide/mm/damon/usage.rst @@ -134,7 +134,8 @@ Users can write below commands for the kdamond to the ``state`` file. - ``on``: Start running. - ``off``: Stop running. - ``commit``: Read the user inputs in the sysfs files except ``state`` file - again. + again. Monitoring :ref:`target region ` inputs are also be + ignored if no target region is specified. - ``update_tuned_intervals``: Update the contents of ``sample_us`` and ``aggr_us`` files of the kdamond with the auto-tuning applied ``sampling interval`` and ``aggregation interval`` for the files. Please refer to @@ -296,6 +297,11 @@ In the beginning, this directory has only one file, ``nr_regions``. Writing a number (``N``) to the file creates the number of child directories named ``0`` to ``N-1``. Each directory represents each initial monitoring target region. +If ``nr_regions`` is zero when committing new DAMON parameters online (writing +``commit`` to ``state`` file of :ref:`kdamond `), the commit +logic ignores the target regions. In other words, the current monitoring +results for the target are preserved. + .. _sysfs_region: regions// -- 2.47.3 Commit 7db551fcfb2a ("mm/damon/reclaim: support addr_unit for DAMON_RECLAIM") introduced the 'addr_unit' parameter for DAMON_RECLAIM. But the usage document is not updated for that. Update the document. Signed-off-by: SeongJae Park --- .../admin-guide/mm/damon/reclaim.rst | 22 +++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/Documentation/admin-guide/mm/damon/reclaim.rst b/Documentation/admin-guide/mm/damon/reclaim.rst index 92bb7cf1b558..552a7786b082 100644 --- a/Documentation/admin-guide/mm/damon/reclaim.rst +++ b/Documentation/admin-guide/mm/damon/reclaim.rst @@ -232,6 +232,28 @@ The end physical address of memory region that DAMON_RECLAIM will do work against. That is, DAMON_RECLAIM will find cold memory regions in this region and reclaims. By default, biggest System RAM is used as the region. +addr_unit +--------- + +A scale factor for memory addresses and bytes. + +This parameter is for setting and getting the :ref:`address unit +` parameter of the DAMON instance for DAMON_RECLAIM. + +``monitor_region_start`` and ``monitor_region_end`` should be provided in this +unit. For example, let's suppose ``addr_unit``, ``monitor_region_start`` and +``monitor_region_end`` are set as ``1024``, ``0`` and ``10``, respectively. +Then DAMON_RECLAIM will work for 10 KiB length of physical address range that +starts from address zero (``[0 * 1024, 10 * 1024)`` in bytes). + +``bytes_reclaim_tried_regions`` and ``bytes_reclaimed_regions`` are also in +this unit. For example, let's suppose values of ``addr_unit``, +``bytes_reclaim_tried_regions`` and ``bytes_reclaimed_regions`` are ``1024``, +``42``, and ``32``, respectively. Then it means DAMON_RECLAIM tried to reclaim +42 KiB memory and successfully reclaimed 32 KiB memory in total. + +If unsure, use only the default value (``1``) and forget about this. + skip_anon --------- -- 2.47.3 Commit 2e0fe9245d6b ("mm/damon/lru_sort: support addr_unit for DAMON_LRU_SORT") introduced the 'addr_unit' parameter for DAMON_LRU_SORT. But the usage document is not updated for that. Update the document. Signed-off-by: SeongJae Park --- .../admin-guide/mm/damon/lru_sort.rst | 22 +++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/Documentation/admin-guide/mm/damon/lru_sort.rst b/Documentation/admin-guide/mm/damon/lru_sort.rst index 7b0775d281b4..72a943202676 100644 --- a/Documentation/admin-guide/mm/damon/lru_sort.rst +++ b/Documentation/admin-guide/mm/damon/lru_sort.rst @@ -211,6 +211,28 @@ End of target memory region in physical address. The end physical address of memory region that DAMON_LRU_SORT will do work against. By default, biggest System RAM is used as the region. +addr_unit +--------- + +A scale factor for memory addresses and bytes. + +This parameter is for setting and getting the :ref:`address unit +` parameter of the DAMON instance for DAMON_RECLAIM. + +``monitor_region_start`` and ``monitor_region_end`` should be provided in this +unit. For example, let's suppose ``addr_unit``, ``monitor_region_start`` and +``monitor_region_end`` are set as ``1024``, ``0`` and ``10``, respectively. +Then DAMON_LRU_SORT will work for 10 KiB length of physical address range that +starts from address zero (``[0 * 1024, 10 * 1024)`` in bytes). + +Stat parameters having ``bytes_`` prefix are also in this unit. For example, +let's suppose values of ``addr_unit``, ``bytes_lru_sort_tried_hot_regions`` and +``bytes_lru_sorted_hot_regions`` are ``1024``, ``42``, and ``32``, +respectively. Then it means DAMON_LRU_SORT tried to LRU-sort 42 KiB of hot +memory and successfully LRU-sorted 32 KiB of the memory in total. + +If unsure, use only the default value (``1``) and forget about this. + kdamond_pid ----------- -- 2.47.3 Commit cc7ceb1d14b0 ("mm/damon/stat: expose the current tuned aggregation interval"), has introduced 'aggr_interval_us' parameter for DAMON_STAT. But the new parameter is not yet documented. Document it on the usage document for the module. Signed-off-by: SeongJae Park --- Documentation/admin-guide/mm/damon/stat.rst | 18 +++++++++++++++++- 1 file changed, 17 insertions(+), 1 deletion(-) diff --git a/Documentation/admin-guide/mm/damon/stat.rst b/Documentation/admin-guide/mm/damon/stat.rst index 20f540a9d3d2..754f98d47617 100644 --- a/Documentation/admin-guide/mm/damon/stat.rst +++ b/Documentation/admin-guide/mm/damon/stat.rst @@ -10,6 +10,8 @@ on the system's entire physical memory using DAMON, and provides simplified access monitoring results statistics, namely idle time percentiles and estimated memory bandwidth. +.. _damon_stat_monitoring_accuracy_overhead: + Monitoring Accuracy and Overhead ================================ @@ -19,7 +21,9 @@ overhead minimum. It auto-tunes the intervals aiming 4 % of observable access events to be captured in each snapshot, while limiting the resulting sampling interval to be 5 milliseconds in minimum and 10 seconds in maximum. On a few production server systems, it resulted in consuming only 0.x % single CPU time, -while capturing reasonable quality of access patterns. +while capturing reasonable quality of access patterns. The tuning-resulting +intervals can be retrieved via ``aggr_interval_us`` :ref:`parameter +`. Interface: Module Parameters ============================ @@ -41,6 +45,18 @@ You can enable DAMON_STAT by setting the value of this parameter as ``Y``. Setting it as ``N`` disables DAMON_STAT. The default value is set by ``CONFIG_DAMON_STAT_ENABLED_DEFAULT`` build config option. +.. _damon_stat_aggr_interval_us: + +aggr_interval_us +---------------- + +Auto-tuned aggregation time interval in microseconds. + +Users can read the aggregation interval of DAMON that is being used by the +DAMON instance for DAMON_STAT. It is :ref:`auto-tuned +` and therefore the value is +dynamically changed. + estimated_memory_bandwidth -------------------------- -- 2.47.3 Commit a983a26d5298 ("mm/damon/stat: expose negative idle time") introduced the negative idle time feature for DAMON_STAT. But it is not documented. Document it on the usage document. Signed-off-by: SeongJae Park --- Documentation/admin-guide/mm/damon/stat.rst | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) diff --git a/Documentation/admin-guide/mm/damon/stat.rst b/Documentation/admin-guide/mm/damon/stat.rst index 754f98d47617..e5a5a2c4f803 100644 --- a/Documentation/admin-guide/mm/damon/stat.rst +++ b/Documentation/admin-guide/mm/damon/stat.rst @@ -74,12 +74,13 @@ memory_idle_ms_percentiles Per-byte idle time (milliseconds) percentiles of the system. DAMON_STAT calculates how long each byte of the memory was not accessed until -now (idle time), based on the current DAMON results snapshot. If DAMON found a -region of access frequency (nr_accesses) larger than zero, every byte of the -region gets zero idle time. If a region has zero access frequency -(nr_accesses), how long the region was keeping the zero access frequency (age) -becomes the idle time of every byte of the region. Then, DAMON_STAT exposes -the percentiles of the idle time values via this read-only parameter. Reading -the parameter returns 101 idle time values in milliseconds, separated by comma. +now (idle time), based on the current DAMON results snapshot. For regions +having access frequency (nr_accesses) larger than zero, how long the current +access frequency level was kept multiplied by ``-1`` becomes the idlee time of +every byte of the region. If a region has zero access frequency (nr_accesses), +how long the region was keeping the zero access frequency (age) becomes the +idle time of every byte of the region. Then, DAMON_STAT exposes the +percentiles of the idle time values via this read-only parameter. Reading the +parameter returns 101 idle time values in milliseconds, separated by comma. Each value represents 0-th, 1st, 2nd, 3rd, ..., 99th and 100th percentile idle times. -- 2.47.3