From 0e2e63dd608bf5844ffae7bf7d860de18a62724c Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 20 May 2010 14:45:26 +0200 Subject: [PATCH] perf-record: Share per-cpu buffers It seems a waste of space to create a buffer per event, share it per-cpu. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker Cc: Paul Mackerras Cc: Mike Galbraith Cc: Steven Rostedt LKML-Reference: <20100521090710.634824884@chello.nl> Signed-off-by: Ingo Molnar --- tools/perf/builtin-record.c | 52 ++++++++++++++++++++----------------- 1 file changed, 28 insertions(+), 24 deletions(-) diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index 94e210f86077..9bc89050e6f8 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -82,7 +82,7 @@ struct mmap_data { unsigned int prev; }; -static struct mmap_data *mmap_array[MAX_NR_CPUS][MAX_COUNTERS]; +static struct mmap_data mmap_array[MAX_NR_CPUS]; static unsigned long mmap_read_head(struct mmap_data *md) { @@ -365,18 +365,29 @@ static void create_counter(int counter, int cpu) if (group && group_fd == -1) group_fd = fd[nr_cpu][counter][thread_index]; - event_array[nr_poll].fd = fd[nr_cpu][counter][thread_index]; - event_array[nr_poll].events = POLLIN; - nr_poll++; - - mmap_array[nr_cpu][counter][thread_index].counter = counter; - mmap_array[nr_cpu][counter][thread_index].prev = 0; - mmap_array[nr_cpu][counter][thread_index].mask = mmap_pages*page_size - 1; - mmap_array[nr_cpu][counter][thread_index].base = mmap(NULL, (mmap_pages+1)*page_size, - PROT_READ|PROT_WRITE, MAP_SHARED, fd[nr_cpu][counter][thread_index], 0); - if (mmap_array[nr_cpu][counter][thread_index].base == MAP_FAILED) { - error("failed to mmap with %d (%s)\n", errno, strerror(errno)); - exit(-1); + if (counter || thread_index) { + ret = ioctl(fd[nr_cpu][counter][thread_index], + PERF_EVENT_IOC_SET_OUTPUT, + fd[nr_cpu][0][0]); + if (ret) { + error("failed to set output: %d (%s)\n", errno, + strerror(errno)); + exit(-1); + } + } else { + mmap_array[nr_cpu].counter = counter; + mmap_array[nr_cpu].prev = 0; + mmap_array[nr_cpu].mask = mmap_pages*page_size - 1; + mmap_array[nr_cpu].base = mmap(NULL, (mmap_pages+1)*page_size, + PROT_READ|PROT_WRITE, MAP_SHARED, fd[nr_cpu][counter][thread_index], 0); + if (mmap_array[nr_cpu].base == MAP_FAILED) { + error("failed to mmap with %d (%s)\n", errno, strerror(errno)); + exit(-1); + } + + event_array[nr_poll].fd = fd[nr_cpu][counter][thread_index]; + event_array[nr_poll].events = POLLIN; + nr_poll++; } if (filter != NULL) { @@ -477,16 +488,11 @@ static struct perf_event_header finished_round_event = { static void mmap_read_all(void) { - int i, counter, thread; + int i; for (i = 0; i < nr_cpu; i++) { - for (counter = 0; counter < nr_counters; counter++) { - for (thread = 0; thread < thread_num; thread++) { - if (mmap_array[i][counter][thread].base) - mmap_read(&mmap_array[i][counter][thread]); - } - - } + if (mmap_array[i].base) + mmap_read(&mmap_array[i]); } if (perf_header__has_feat(&session->header, HEADER_TRACE_INFO)) @@ -861,9 +867,7 @@ int cmd_record(int argc, const char **argv, const char *prefix __used) for (i = 0; i < MAX_NR_CPUS; i++) { for (j = 0; j < MAX_COUNTERS; j++) { fd[i][j] = malloc(sizeof(int)*thread_num); - mmap_array[i][j] = zalloc( - sizeof(struct mmap_data)*thread_num); - if (!fd[i][j] || !mmap_array[i][j]) + if (!fd[i][j]) return -ENOMEM; } } -- GitLab