From 54357f0c9149c871e5e4b83ad385a6f2ad3a749f Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 10 Aug 2021 15:26:25 +0200 Subject: [PATCH 01/13] tracing: Add migrate-disabled counter to tracing output. migrate_disable() forbids task migration to another CPU. It is available since v5.11 and has already users such as highmem or BPF. It is useful to observe this task state in tracing which already has other states like the preemption counter. Instead of adding the migrate disable counter as a new entry to struct trace_entry, which would extend the whole struct by four bytes, it is squashed into the preempt-disable counter. The lower four bits represent the preemption counter, the upper four bits represent the migrate disable counter. Both counter shouldn't exceed 15 but if they do, there is a safety net which caps the value at 15. Add the migrate-disable counter to the trace entry so it shows up in the trace. Due to the users mentioned above, it is already possible to observe it: | bash-1108 [000] ...21 73.950578: rss_stat: mm_id=2213312838 curr=0 type=MM_ANONPAGES size=8192B | bash-1108 [000] d..31 73.951222: irq_disable: caller=flush_tlb_mm_range+0x115/0x130 parent=ptep_clear_flush+0x42/0x50 | bash-1108 [000] d..31 73.951222: tlb_flush: pages:1 reason:local mm shootdown (3) The last value is the migrate-disable counter. Things that popped up: - trace_print_lat_context() does not print the migrate counter. Not sure if it should. It is used in "verbose" mode and uses 8 digits and I'm not sure ther is something processing the value. - trace_define_common_fields() now defines a different variable. This probably breaks things. No ide what to do in order to preserve the old behaviour. Since this is used as a filter it should be split somehow to be able to match both nibbles here. Link: https://lkml.kernel.org/r/20210810132625.ylssabmsrkygokuv@linutronix.de Signed-off-by: Thomas Gleixner [bigeasy: patch description.] Signed-off-by: Sebastian Andrzej Siewior [ SDR: Removed change to common_preempt_count field name ] Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace.c | 26 +++++++++++++++++++------- kernel/trace/trace_events.c | 1 + kernel/trace/trace_output.c | 11 ++++++++--- 3 files changed, 28 insertions(+), 10 deletions(-) diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 489924cde4f8..b554e18924f8 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -2603,6 +2603,15 @@ enum print_line_t trace_handle_return(struct trace_seq *s) } EXPORT_SYMBOL_GPL(trace_handle_return); +static unsigned short migration_disable_value(void) +{ +#if defined(CONFIG_SMP) + return current->migration_disabled; +#else + return 0; +#endif +} + unsigned int tracing_gen_ctx_irq_test(unsigned int irqs_status) { unsigned int trace_flags = irqs_status; @@ -2621,7 +2630,8 @@ unsigned int tracing_gen_ctx_irq_test(unsigned int irqs_status) trace_flags |= TRACE_FLAG_NEED_RESCHED; if (test_preempt_need_resched()) trace_flags |= TRACE_FLAG_PREEMPT_RESCHED; - return (trace_flags << 16) | (pc & 0xff); + return (trace_flags << 16) | (min_t(unsigned int, pc & 0xff, 0xf)) | + (min_t(unsigned int, migration_disable_value(), 0xf)) << 4; } struct ring_buffer_event * @@ -4189,9 +4199,10 @@ static void print_lat_help_header(struct seq_file *m) "# | / _----=> need-resched \n" "# || / _---=> hardirq/softirq \n" "# ||| / _--=> preempt-depth \n" - "# |||| / delay \n" - "# cmd pid ||||| time | caller \n" - "# \\ / ||||| \\ | / \n"); + "# |||| / _-=> migrate-disable \n" + "# ||||| / delay \n" + "# cmd pid |||||| time | caller \n" + "# \\ / |||||| \\ | / \n"); } static void print_event_info(struct array_buffer *buf, struct seq_file *m) @@ -4229,9 +4240,10 @@ static void print_func_help_header_irq(struct array_buffer *buf, struct seq_file seq_printf(m, "# %.*s / _----=> need-resched\n", prec, space); seq_printf(m, "# %.*s| / _---=> hardirq/softirq\n", prec, space); seq_printf(m, "# %.*s|| / _--=> preempt-depth\n", prec, space); - seq_printf(m, "# %.*s||| / delay\n", prec, space); - seq_printf(m, "# TASK-PID %.*s CPU# |||| TIMESTAMP FUNCTION\n", prec, " TGID "); - seq_printf(m, "# | | %.*s | |||| | |\n", prec, " | "); + seq_printf(m, "# %.*s||| / _-=> migrate-disable\n", prec, space); + seq_printf(m, "# %.*s|||| / delay\n", prec, space); + seq_printf(m, "# TASK-PID %.*s CPU# ||||| TIMESTAMP FUNCTION\n", prec, " TGID "); + seq_printf(m, "# | | %.*s | ||||| | |\n", prec, " | "); } void diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 1349b6de5eeb..830b3b9940f4 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -181,6 +181,7 @@ static int trace_define_common_fields(void) __common_field(unsigned short, type); __common_field(unsigned char, flags); + /* Holds both preempt_count and migrate_disable */ __common_field(unsigned char, preempt_count); __common_field(int, pid); diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c index a0bf446bb034..c2ca40e8595b 100644 --- a/kernel/trace/trace_output.c +++ b/kernel/trace/trace_output.c @@ -492,8 +492,13 @@ int trace_print_lat_fmt(struct trace_seq *s, struct trace_entry *entry) trace_seq_printf(s, "%c%c%c", irqs_off, need_resched, hardsoft_irq); - if (entry->preempt_count) - trace_seq_printf(s, "%x", entry->preempt_count); + if (entry->preempt_count & 0xf) + trace_seq_printf(s, "%x", entry->preempt_count & 0xf); + else + trace_seq_putc(s, '.'); + + if (entry->preempt_count & 0xf0) + trace_seq_printf(s, "%x", entry->preempt_count >> 4); else trace_seq_putc(s, '.'); @@ -656,7 +661,7 @@ int trace_print_lat_context(struct trace_iterator *iter) trace_seq_printf( s, "%16s %7d %3d %d %08x %08lx ", comm, entry->pid, iter->cpu, entry->flags, - entry->preempt_count, iter->idx); + entry->preempt_count & 0xf, iter->idx); } else { lat_print_generic(s, entry, iter->cpu); } From 5615e088b43d564aec08ccfaecb21ba484a1b4d6 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Tue, 24 Aug 2021 14:51:50 +0300 Subject: [PATCH 02/13] tracing: Fix some alloc_event_probe() error handling bugs There are two bugs in this code. First, if the kzalloc() fails it leads to a NULL dereference of "ep" on the next line. Second, if the alloc_event_probe() function returns an error then it leads to an error pointer dereference in the caller. Link: https://lkml.kernel.org/r/20210824115150.GI31143@kili Fixes: 7491e2c44278 ("tracing: Add a probe that attaches to trace events") Signed-off-by: Dan Carpenter Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_eprobe.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/kernel/trace/trace_eprobe.c b/kernel/trace/trace_eprobe.c index 56a96e9750cf..3044b762cbd7 100644 --- a/kernel/trace/trace_eprobe.c +++ b/kernel/trace/trace_eprobe.c @@ -151,7 +151,7 @@ static struct trace_eprobe *alloc_event_probe(const char *group, ep = kzalloc(struct_size(ep, tp.args, nargs), GFP_KERNEL); if (!ep) { - trace_event_put_ref(ep->event); + trace_event_put_ref(event); goto error; } ep->event = event; @@ -851,7 +851,8 @@ static int __trace_eprobe_create(int argc, const char *argv[]) ret = PTR_ERR(ep); /* This must return -ENOMEM, else there is a bug */ WARN_ON_ONCE(ret != -ENOMEM); - goto error; /* We know ep is not allocated */ + ep = NULL; + goto error; } argc -= 2; argv += 2; From 4b6b08f2e45edda4c067ac40833e3c1f84383c0b Mon Sep 17 00:00:00 2001 From: "Qiang.Zhang" Date: Tue, 31 Aug 2021 10:29:19 +0800 Subject: [PATCH 03/13] tracing/osnoise: Fix missed cpus_read_unlock() in start_per_cpu_kthreads() When start_kthread() return error, the cpus_read_unlock() need to be called. Link: https://lkml.kernel.org/r/20210831022919.27630-1-qiang.zhang@windriver.com Cc: Fixes: c8895e271f79 ("trace/osnoise: Support hotplug operations") Acked-by: Daniel Bristot de Oliveira Signed-off-by: Qiang.Zhang Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_osnoise.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/kernel/trace/trace_osnoise.c b/kernel/trace/trace_osnoise.c index 65b08b8e5bf8..ce053619f289 100644 --- a/kernel/trace/trace_osnoise.c +++ b/kernel/trace/trace_osnoise.c @@ -1548,7 +1548,7 @@ static int start_kthread(unsigned int cpu) static int start_per_cpu_kthreads(struct trace_array *tr) { struct cpumask *current_mask = &save_cpumask; - int retval; + int retval = 0; int cpu; cpus_read_lock(); @@ -1568,13 +1568,13 @@ static int start_per_cpu_kthreads(struct trace_array *tr) retval = start_kthread(cpu); if (retval) { stop_per_cpu_kthreads(); - return retval; + break; } } cpus_read_unlock(); - return 0; + return retval; } #ifdef CONFIG_HOTPLUG_CPU From 40caa127f3c7279c75cb0c9684559fa314ee3a66 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Sun, 5 Sep 2021 00:54:09 +0900 Subject: [PATCH 04/13] init: bootconfig: Remove all bootconfig data when the init memory is removed Since the bootconfig is used only in the init functions, it doesn't need to keep the data after boot. Free it when the init memory is removed. Link: https://lkml.kernel.org/r/163077084958.222577.5924961258513004428.stgit@devnote2 Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- init/main.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/init/main.c b/init/main.c index 8d97aba78c3a..d35c4a865adb 100644 --- a/init/main.c +++ b/init/main.c @@ -468,7 +468,12 @@ static void __init setup_boot_config(void) return; } -#else +static void __init exit_boot_config(void) +{ + xbc_destroy_all(); +} + +#else /* !CONFIG_BOOT_CONFIG */ static void __init setup_boot_config(void) { @@ -481,7 +486,11 @@ static int __init warn_bootconfig(char *str) pr_warn("WARNING: 'bootconfig' found on the kernel command line but CONFIG_BOOT_CONFIG is not set.\n"); return 0; } -#endif + +#define exit_boot_config() do {} while (0) + +#endif /* CONFIG_BOOT_CONFIG */ + early_param("bootconfig", warn_bootconfig); /* Change NUL term back to "=", to make "param" the whole string. */ @@ -1493,6 +1502,7 @@ static int __ref kernel_init(void *unused) kprobe_free_init_mem(); ftrace_free_init_mem(); kgdb_free_init_mem(); + exit_boot_config(); free_initmem(); mark_readonly(); From b66fbbe8d48228e1e81b583ae39abb86d7d09053 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Sun, 5 Sep 2021 00:54:16 +0900 Subject: [PATCH 05/13] init/bootconfig: Reorder init parameter from bootconfig and cmdline Reorder the init parameters from bootconfig and kernel cmdline so that the kernel cmdline always be the last part of the parameters as below. " -- "[bootconfig init params][cmdline init params] This change will help us to prevent that bootconfig init params overwrite the init params which user gives in the command line. Link: https://lkml.kernel.org/r/163077085675.222577.5665176468023636160.stgit@devnote2 Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- init/main.c | 23 ++++++++++++++--------- 1 file changed, 14 insertions(+), 9 deletions(-) diff --git a/init/main.c b/init/main.c index d35c4a865adb..d08caed17c7f 100644 --- a/init/main.c +++ b/init/main.c @@ -153,10 +153,10 @@ static char *extra_init_args; #ifdef CONFIG_BOOT_CONFIG /* Is bootconfig on command line? */ static bool bootconfig_found; -static bool initargs_found; +static size_t initargs_offs; #else # define bootconfig_found false -# define initargs_found false +# define initargs_offs 0 #endif static char *execute_command; @@ -422,9 +422,9 @@ static void __init setup_boot_config(void) if (IS_ERR(err) || !bootconfig_found) return; - /* parse_args() stops at '--' and returns an address */ + /* parse_args() stops at the next param of '--' and returns an address */ if (err) - initargs_found = true; + initargs_offs = err - tmp_cmdline; if (!data) { pr_err("'bootconfig' found on command line, but no bootconfig found\n"); @@ -655,16 +655,21 @@ static void __init setup_command_line(char *command_line) * Append supplemental init boot args to saved_command_line * so that user can check what command line options passed * to init. + * The order should always be + * " -- "[bootconfig init-param][cmdline init-param] */ - len = strlen(saved_command_line); - if (initargs_found) { - saved_command_line[len++] = ' '; + if (initargs_offs) { + len = xlen + initargs_offs; + strcpy(saved_command_line + len, extra_init_args); + len += ilen - 4; /* strlen(extra_init_args) */ + strcpy(saved_command_line + len, + boot_command_line + initargs_offs - 1); } else { + len = strlen(saved_command_line); strcpy(saved_command_line + len, " -- "); len += 4; + strcpy(saved_command_line + len, extra_init_args); } - - strcpy(saved_command_line + len, extra_init_args); } } From 26c9c72fd0b960906b6a31a614fb30d2434b166b Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Sun, 5 Sep 2021 00:54:24 +0900 Subject: [PATCH 06/13] docs: bootconfig: Add how to use bootconfig for kernel parameters Add a section to describe how to use the bootconfig for specifying kernel and init parameters. This is an important section because it is the reason why this document is under the admin-guide. Link: https://lkml.kernel.org/r/163077086399.222577.5881779375643977991.stgit@devnote2 Signed-off-by: Masami Hiramatsu Cc: Jonathan Corbet Cc: linux-doc@vger.kernel.org Signed-off-by: Steven Rostedt (VMware) --- Documentation/admin-guide/bootconfig.rst | 39 +++++++++++++++++++++++- 1 file changed, 38 insertions(+), 1 deletion(-) diff --git a/Documentation/admin-guide/bootconfig.rst b/Documentation/admin-guide/bootconfig.rst index 6a79f2e59396..a1860fc0ca88 100644 --- a/Documentation/admin-guide/bootconfig.rst +++ b/Documentation/admin-guide/bootconfig.rst @@ -178,7 +178,7 @@ update the boot loader and the kernel image itself as long as the boot loader passes the correct initrd file size. If by any chance, the boot loader passes a longer size, the kernel fails to find the bootconfig data. -To do this operation, Linux kernel provides "bootconfig" command under +To do this operation, Linux kernel provides ``bootconfig`` command under tools/bootconfig, which allows admin to apply or delete the config file to/from initrd image. You can build it by the following command:: @@ -196,6 +196,43 @@ To remove the config from the image, you can use -d option as below:: Then add "bootconfig" on the normal kernel command line to tell the kernel to look for the bootconfig at the end of the initrd file. + +Kernel parameters via Boot Config +================================= + +In addition to the kernel command line, the boot config can be used for +passing the kernel parameters. All the key-value pairs under ``kernel`` +key will be passed to kernel cmdline directly. Moreover, the key-value +pairs under ``init`` will be passed to init process via the cmdline. +The parameters are concatinated with user-given kernel cmdline string +as the following order, so that the command line parameter can override +bootconfig parameters (this depends on how the subsystem handles parameters +but in general, earlier parameter will be overwritten by later one.):: + + [bootconfig params][cmdline params] -- [bootconfig init params][cmdline init params] + +Here is an example of the bootconfig file for kernel/init parameters.:: + + kernel { + root = 01234567-89ab-cdef-0123-456789abcd + } + init { + splash + } + +This will be copied into the kernel cmdline string as the following:: + + root="01234567-89ab-cdef-0123-456789abcd" -- splash + +If user gives some other command line like,:: + + ro bootconfig -- quiet + +The final kernel cmdline will be the following:: + + root="01234567-89ab-cdef-0123-456789abcd" ro bootconfig -- splash quiet + + Config File Limitation ====================== From 32ba9f0fb027cc43074e3ea26fcf831adeee8e03 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Sun, 5 Sep 2021 00:54:31 +0900 Subject: [PATCH 07/13] tools/bootconfig: Fix tracing_on option checking in ftrace2bconf.sh Since tracing_on indicates only "1" (default) or "0", ftrace2bconf.sh only need to check the value is "0". Link: https://lkml.kernel.org/r/163077087144.222577.6888011847727968737.stgit@devnote2 Fixes: 55ed4560774d ("tools/bootconfig: Add tracing_on support to helper scripts") Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- tools/bootconfig/scripts/ftrace2bconf.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/bootconfig/scripts/ftrace2bconf.sh b/tools/bootconfig/scripts/ftrace2bconf.sh index fbaf07dc91bf..6183b36c6846 100755 --- a/tools/bootconfig/scripts/ftrace2bconf.sh +++ b/tools/bootconfig/scripts/ftrace2bconf.sh @@ -239,8 +239,8 @@ instance_options() { # [instance-name] emit_kv $PREFIX.cpumask = $val fi val=`cat $INSTANCE/tracing_on` - if [ `echo $val | sed -e s/f//g`x != x ]; then - emit_kv $PREFIX.tracing_on = $val + if [ "$val" = "0" ]; then + emit_kv $PREFIX.tracing_on = 0 fi val=`cat $INSTANCE/current_tracer` From 903bd067faa837fddb6e5c8b740c3374dc582f04 Mon Sep 17 00:00:00 2001 From: Julio Faracco Date: Sun, 5 Sep 2021 00:54:38 +0900 Subject: [PATCH 08/13] bootconfig: Fix missing return check of xbc_node_compose_key function The function `xbc_show_list should` handle the keys during the composition. Even the errors returned by the compose function. Instead of removing the `ret` variable, it should save the value and show the exact error. This missing variable is causing a compilation issue also. Link: https://lkml.kernel.org/r/163077087861.222577.12884543474750968146.stgit@devnote2 Fixes: e5efaeb8a8f5 ("bootconfig: Support mixing a value and subkeys under a key") Signed-off-by: Julio Faracco Acked-by: Masami Hiramatsu Cc: stable@vger.kernel.org Signed-off-by: Steven Rostedt (VMware) --- tools/bootconfig/main.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/tools/bootconfig/main.c b/tools/bootconfig/main.c index f45fa992e01d..fd67496a947f 100644 --- a/tools/bootconfig/main.c +++ b/tools/bootconfig/main.c @@ -111,9 +111,11 @@ static void xbc_show_list(void) char key[XBC_KEYLEN_MAX]; struct xbc_node *leaf; const char *val; + int ret; xbc_for_each_key_value(leaf, val) { - if (xbc_node_compose_key(leaf, key, XBC_KEYLEN_MAX) < 0) { + ret = xbc_node_compose_key(leaf, key, XBC_KEYLEN_MAX); + if (ret < 0) { fprintf(stderr, "Failed to compose key %d\n", ret); break; } From 47914d4e591cbbf7ce5103d86257a5d0d8532110 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Sun, 5 Sep 2021 00:54:46 +0900 Subject: [PATCH 09/13] tools/bootconfig: Show whole test command for each test case Show whole test command instead of only the 3rd argument. This will clear to show what will be actually tested by each test case. Link: https://lkml.kernel.org/r/163077088607.222577.14786016266462495017.stgit@devnote2 Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- tools/bootconfig/test-bootconfig.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/bootconfig/test-bootconfig.sh b/tools/bootconfig/test-bootconfig.sh index baed891d0ba4..f68e2e9eef8b 100755 --- a/tools/bootconfig/test-bootconfig.sh +++ b/tools/bootconfig/test-bootconfig.sh @@ -26,7 +26,7 @@ trap cleanup EXIT TERM NO=1 xpass() { # pass test command - echo "test case $NO ($3)... " + echo "test case $NO ($*)... " if ! ($@ && echo "\t\t[OK]"); then echo "\t\t[NG]"; NG=$((NG + 1)) fi @@ -34,7 +34,7 @@ xpass() { # pass test command } xfail() { # fail test command - echo "test case $NO ($3)... " + echo "test case $NO ($*)... " if ! (! $@ && echo "\t\t[OK]"); then echo "\t\t[NG]"; NG=$((NG + 1)) fi From 0be083cee42ec78ba6f9148f5b12a00aa2fb8bd3 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Wed, 1 Sep 2021 16:55:13 +0300 Subject: [PATCH 10/13] tracing: synth events: increase max fields count Sometimes it is useful to construct larger synthetic trace events. Increase 'SYNTH_FIELDS_MAX' (maximum number of fields in a synthetic event) from 32 to 64. Link: https://lkml.kernel.org/r/20210901135513.3087062-1-dedekind1@gmail.com Signed-off-by: Artem Bityutskiy Acked-by: Tom Zanussi Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_synth.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/trace/trace_synth.h b/kernel/trace/trace_synth.h index 4007fe95cf42..b29595fe3ac5 100644 --- a/kernel/trace/trace_synth.h +++ b/kernel/trace/trace_synth.h @@ -5,7 +5,7 @@ #include "trace_dynevent.h" #define SYNTH_SYSTEM "synthetic" -#define SYNTH_FIELDS_MAX 32 +#define SYNTH_FIELDS_MAX 64 #define STR_VAR_LEN_MAX MAX_FILTER_STR_VAL /* must be multiple of sizeof(u64) */ From c910db943d35d4ac4b77570ece76e0799af24233 Mon Sep 17 00:00:00 2001 From: Tom Zanussi Date: Thu, 2 Sep 2021 15:57:12 -0500 Subject: [PATCH 11/13] tracing: Dynamically allocate the per-elt hist_elt_data array Setting the hist_elt_data.field_var_str[] array unconditionally to a size of SYNTH_FIELD_MAX elements wastes space unnecessarily. The actual number of elements needed can be calculated at run-time instead. In most cases, this will save a lot of space since it's a per-elt array which isn't normally close to being full. It also allows us to increase SYNTH_FIELD_MAX without worrying about even more wastage when we do that. Link: https://lkml.kernel.org/r/d52ae0ad5e1b59af7c4f54faf3fc098461fd82b3.camel@kernel.org Signed-off-by: Tom Zanussi Tested-by: Artem Bityutskiy Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_events_hist.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/kernel/trace/trace_events_hist.c b/kernel/trace/trace_events_hist.c index 9d91b1c06957..a6061a69aa84 100644 --- a/kernel/trace/trace_events_hist.c +++ b/kernel/trace/trace_events_hist.c @@ -508,7 +508,8 @@ struct track_data { struct hist_elt_data { char *comm; u64 *var_ref_vals; - char *field_var_str[SYNTH_FIELDS_MAX]; + char **field_var_str; + int n_field_var_str; }; struct snapshot_context { @@ -1401,9 +1402,11 @@ static void hist_elt_data_free(struct hist_elt_data *elt_data) { unsigned int i; - for (i = 0; i < SYNTH_FIELDS_MAX; i++) + for (i = 0; i < elt_data->n_field_var_str; i++) kfree(elt_data->field_var_str[i]); + kfree(elt_data->field_var_str); + kfree(elt_data->comm); kfree(elt_data); } @@ -1451,6 +1454,13 @@ static int hist_trigger_elt_data_alloc(struct tracing_map_elt *elt) size = STR_VAR_LEN_MAX; + elt_data->field_var_str = kcalloc(n_str, sizeof(char *), GFP_KERNEL); + if (!elt_data->field_var_str) { + hist_elt_data_free(elt_data); + return -EINVAL; + } + elt_data->n_field_var_str = n_str; + for (i = 0; i < n_str; i++) { elt_data->field_var_str[i] = kzalloc(size, GFP_KERNEL); if (!elt_data->field_var_str[i]) { From 04178ea130a64a50826f17c7cc92774ada7ea9d0 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Tue, 7 Sep 2021 23:04:29 -0400 Subject: [PATCH 12/13] selftests/ftrace: Exclude "(fault)" in testing add/remove eprobe events The original test for adding and removing eprobes used synthetic events and retrieved the filename from the open system call at the end of the system call. This would allow it to always be loaded into the page tables when accessed. Masami suggested that the test was too complex for just testing add and remove, so it was changed to test just adding and removing an event probe on top of the start of the open system call event. Now it is possible that the filename will not be loaded into memory at the time the eprobe is triggered, and will result in "(fault)" being displayed in the event. This causes the test to fail. Account for "(fault)" also being one of the values of the filename field of the event probe. Link: https://lkml.kernel.org/r/20210907230429.5783d519@rorschach.local.home Fixes: 079db70794ec5 ("selftests/ftrace: Add test case to test adding and removing of event probe") Acked-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- .../selftests/ftrace/test.d/dynevent/add_remove_eprobe.tc | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/ftrace/test.d/dynevent/add_remove_eprobe.tc b/tools/testing/selftests/ftrace/test.d/dynevent/add_remove_eprobe.tc index 25a3da4eaa44..5f5b2ba3e557 100644 --- a/tools/testing/selftests/ftrace/test.d/dynevent/add_remove_eprobe.tc +++ b/tools/testing/selftests/ftrace/test.d/dynevent/add_remove_eprobe.tc @@ -22,7 +22,7 @@ ls echo 0 > events/eprobes/$EPROBE/enable content=`grep '^ *ls-' trace | grep 'file='` -nocontent=`grep '^ *ls-' trace | grep 'file=' | grep -v -e '"/' -e '"."'` || true +nocontent=`grep '^ *ls-' trace | grep 'file=' | grep -v -e '"/' -e '"."' -e '(fault)' ` || true if [ -z "$content" ]; then exit_fail From cfd799837dbc48499abb05d1891b3d9992354d3a Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Thu, 9 Sep 2021 04:38:03 +0900 Subject: [PATCH 13/13] tracing/boot: Fix to loop on only subkeys Since the commit e5efaeb8a8f5 ("bootconfig: Support mixing a value and subkeys under a key") allows to co-exist a value node and key nodes under a node, xbc_node_for_each_child() is not only returning key node but also a value node. In the boot-time tracing using xbc_node_for_each_child() to iterate the events, groups and instances, but those must be key nodes. Thus it must use xbc_node_for_each_subkey(). Link: https://lkml.kernel.org/r/163112988361.74896.2267026262061819145.stgit@devnote2 Fixes: e5efaeb8a8f5 ("bootconfig: Support mixing a value and subkeys under a key") Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_boot.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/kernel/trace/trace_boot.c b/kernel/trace/trace_boot.c index 1060b0446032..388e65d05978 100644 --- a/kernel/trace/trace_boot.c +++ b/kernel/trace/trace_boot.c @@ -522,14 +522,14 @@ trace_boot_init_events(struct trace_array *tr, struct xbc_node *node) if (!node) return; /* per-event key starts with "event.GROUP.EVENT" */ - xbc_node_for_each_child(node, gnode) { + xbc_node_for_each_subkey(node, gnode) { data = xbc_node_get_data(gnode); if (!strcmp(data, "enable")) { enable_all = true; continue; } enable = false; - xbc_node_for_each_child(gnode, enode) { + xbc_node_for_each_subkey(gnode, enode) { data = xbc_node_get_data(enode); if (!strcmp(data, "enable")) { enable = true; @@ -625,7 +625,7 @@ trace_boot_init_instances(struct xbc_node *node) if (!node) return; - xbc_node_for_each_child(node, inode) { + xbc_node_for_each_subkey(node, inode) { p = xbc_node_get_data(inode); if (!p || *p == '\0') continue;