From c58a3f8c7f974d171d1b6897a71a078a3bc7afd3 Mon Sep 17 00:00:00 2001 From: Daniel Bristot de Oliveira Date: Tue, 6 Jun 2023 18:12:18 +0200 Subject: rtla: Automatically move rtla to a house-keeping cpu When the user sets -c try to move rtla out of the , even without an -H option. This is useful to avoid having rtla interfering with the workload. This works by removing from rtla's current affinity. If rtla fails to move itself away it is not that of a problem as this is an automatic measure. Link: https://lkml.kernel.org/r/c54304d90c777310fb85a3e658d1449173759aab.1686066600.git.bristot@kernel.org Cc: William White Cc: Jonathan Corbet Tested-by: Juri Lelli Signed-off-by: Daniel Bristot de Oliveira Signed-off-by: Steven Rostedt (Google) --- tools/tracing/rtla/src/osnoise_hist.c | 9 ++++++ tools/tracing/rtla/src/osnoise_top.c | 9 ++++++ tools/tracing/rtla/src/timerlat_hist.c | 9 ++++++ tools/tracing/rtla/src/timerlat_top.c | 9 ++++++ tools/tracing/rtla/src/utils.c | 50 ++++++++++++++++++++++++++++++++++ tools/tracing/rtla/src/utils.h | 1 + 6 files changed, 87 insertions(+) (limited to 'tools/tracing') diff --git a/tools/tracing/rtla/src/osnoise_hist.c b/tools/tracing/rtla/src/osnoise_hist.c index b616a72d5c0a..dfbcb5ca7ecb 100644 --- a/tools/tracing/rtla/src/osnoise_hist.c +++ b/tools/tracing/rtla/src/osnoise_hist.c @@ -753,6 +753,15 @@ osnoise_hist_apply_config(struct osnoise_tool *tool, struct osnoise_hist_params err_msg("Failed to set rtla to the house keeping CPUs\n"); goto out_err; } + } else if (params->cpus) { + /* + * Even if the user do not set a house-keeping CPU, try to + * move rtla to a CPU set different to the one where the user + * set the workload to run. + * + * No need to check results as this is an automatic attempt. + */ + auto_house_keeping(¶ms->monitored_cpus); } return 0; diff --git a/tools/tracing/rtla/src/osnoise_top.c b/tools/tracing/rtla/src/osnoise_top.c index 0e2f3b216d34..0833537bb2eb 100644 --- a/tools/tracing/rtla/src/osnoise_top.c +++ b/tools/tracing/rtla/src/osnoise_top.c @@ -582,6 +582,15 @@ osnoise_top_apply_config(struct osnoise_tool *tool, struct osnoise_top_params *p err_msg("Failed to set rtla to the house keeping CPUs\n"); goto out_err; } + } else if (params->cpus) { + /* + * Even if the user do not set a house-keeping CPU, try to + * move rtla to a CPU set different to the one where the user + * set the workload to run. + * + * No need to check results as this is an automatic attempt. + */ + auto_house_keeping(¶ms->monitored_cpus); } return 0; diff --git a/tools/tracing/rtla/src/timerlat_hist.c b/tools/tracing/rtla/src/timerlat_hist.c index 00287e96f22e..1675d54cae81 100644 --- a/tools/tracing/rtla/src/timerlat_hist.c +++ b/tools/tracing/rtla/src/timerlat_hist.c @@ -776,6 +776,15 @@ timerlat_hist_apply_config(struct osnoise_tool *tool, struct timerlat_hist_param err_msg("Failed to set rtla to the house keeping CPUs\n"); goto out_err; } + } else if (params->cpus) { + /* + * Even if the user do not set a house-keeping CPU, try to + * move rtla to a CPU set different to the one where the user + * set the workload to run. + * + * No need to check results as this is an automatic attempt. + */ + auto_house_keeping(¶ms->monitored_cpus); } return 0; diff --git a/tools/tracing/rtla/src/timerlat_top.c b/tools/tracing/rtla/src/timerlat_top.c index 920f2f6ef842..f0c6d9735e2a 100644 --- a/tools/tracing/rtla/src/timerlat_top.c +++ b/tools/tracing/rtla/src/timerlat_top.c @@ -619,6 +619,15 @@ timerlat_top_apply_config(struct osnoise_tool *top, struct timerlat_top_params * err_msg("Failed to set rtla to the house keeping CPUs\n"); goto out_err; } + } else if (params->cpus) { + /* + * Even if the user do not set a house-keeping CPU, try to + * move rtla to a CPU set different to the one where the user + * set the workload to run. + * + * No need to check results as this is an automatic attempt. + */ + auto_house_keeping(¶ms->monitored_cpus); } return 0; diff --git a/tools/tracing/rtla/src/utils.c b/tools/tracing/rtla/src/utils.c index 8f9ad8f01e0f..3e25f0277fb9 100644 --- a/tools/tracing/rtla/src/utils.c +++ b/tools/tracing/rtla/src/utils.c @@ -709,3 +709,53 @@ out_cg: close(cg_fd); return 0; } + +/** + * auto_house_keeping - Automatically move rtla out of measurement threads + * + * Try to move rtla away from the tracer, if possible. + * + * Returns 1 on success, 0 otherwise. + */ +int auto_house_keeping(cpu_set_t *monitored_cpus) +{ + cpu_set_t rtla_cpus, house_keeping_cpus; + int retval; + + /* first get the CPUs in which rtla can actually run. */ + retval = sched_getaffinity(getpid(), sizeof(rtla_cpus), &rtla_cpus); + if (retval == -1) { + debug_msg("Could not get rtla affinity, rtla might run with the threads!\n"); + return 0; + } + + /* then check if the existing setup is already good. */ + CPU_AND(&house_keeping_cpus, &rtla_cpus, monitored_cpus); + if (!CPU_COUNT(&house_keeping_cpus)) { + debug_msg("rtla and the monitored CPUs do not share CPUs."); + debug_msg("Skipping auto house-keeping\n"); + return 1; + } + + /* remove the intersection */ + CPU_XOR(&house_keeping_cpus, &rtla_cpus, monitored_cpus); + + /* get only those that rtla can run */ + CPU_AND(&house_keeping_cpus, &house_keeping_cpus, &rtla_cpus); + + /* is there any cpu left? */ + if (!CPU_COUNT(&house_keeping_cpus)) { + debug_msg("Could not find any CPU for auto house-keeping\n"); + return 0; + } + + retval = sched_setaffinity(getpid(), sizeof(house_keeping_cpus), &house_keeping_cpus); + if (retval == -1) { + debug_msg("Could not set affinity for auto house-keeping\n"); + return 0; + } + + debug_msg("rtla automatically moved to an auto house-keeping cpu set\n"); + + return 1; +} diff --git a/tools/tracing/rtla/src/utils.h b/tools/tracing/rtla/src/utils.h index 9ab2f0d7bc1c..dec59163cfbc 100644 --- a/tools/tracing/rtla/src/utils.h +++ b/tools/tracing/rtla/src/utils.h @@ -60,6 +60,7 @@ int parse_cpu_set(char *cpu_list, cpu_set_t *set); int set_comm_sched_attr(const char *comm_prefix, struct sched_attr *attr); int set_comm_cgroup(const char *comm_prefix, const char *cgroup); int set_cpu_dma_latency(int32_t latency); +int auto_house_keeping(cpu_set_t *monitored_cpus); #define ns_to_usf(x) (((double)x/1000)) #define ns_to_per(total, part) ((part * 100) / (double)total) -- cgit v1.2.3