From 7162431dcf72032835d369c8d7b51311df407938 Mon Sep 17 00:00:00 2001 From: Miroslav Benes Date: Wed, 16 Oct 2019 13:33:13 +0200 Subject: ftrace: Introduce PERMANENT ftrace_ops flag Livepatch uses ftrace for redirection to new patched functions. It means that if ftrace is disabled, all live patched functions are disabled as well. Toggling global 'ftrace_enabled' sysctl thus affect it directly. It is not a problem per se, because only administrator can set sysctl values, but it still may be surprising. Introduce PERMANENT ftrace_ops flag to amend this. If the FTRACE_OPS_FL_PERMANENT is set on any ftrace ops, the tracing cannot be disabled by disabling ftrace_enabled. Equally, a callback with the flag set cannot be registered if ftrace_enabled is disabled. Link: http://lkml.kernel.org/r/20191016113316.13415-2-mbenes@suse.cz Reviewed-by: Petr Mladek Reviewed-by: Kamalesh Babulal Signed-off-by: Miroslav Benes Signed-off-by: Steven Rostedt (VMware) --- kernel/livepatch/patch.c | 3 ++- kernel/trace/ftrace.c | 23 +++++++++++++++++++++-- 2 files changed, 23 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/livepatch/patch.c b/kernel/livepatch/patch.c index bd43537702bd..b552cf2d85f8 100644 --- a/kernel/livepatch/patch.c +++ b/kernel/livepatch/patch.c @@ -196,7 +196,8 @@ static int klp_patch_func(struct klp_func *func) ops->fops.func = klp_ftrace_handler; ops->fops.flags = FTRACE_OPS_FL_SAVE_REGS | FTRACE_OPS_FL_DYNAMIC | - FTRACE_OPS_FL_IPMODIFY; + FTRACE_OPS_FL_IPMODIFY | + FTRACE_OPS_FL_PERMANENT; list_add(&ops->node, &klp_ops); diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index f296d89be757..89e9128652ef 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -326,6 +326,8 @@ int __register_ftrace_function(struct ftrace_ops *ops) if (ops->flags & FTRACE_OPS_FL_SAVE_REGS_IF_SUPPORTED) ops->flags |= FTRACE_OPS_FL_SAVE_REGS; #endif + if (!ftrace_enabled && (ops->flags & FTRACE_OPS_FL_PERMANENT)) + return -EBUSY; if (!core_kernel_data((unsigned long)ops)) ops->flags |= FTRACE_OPS_FL_DYNAMIC; @@ -6754,6 +6756,18 @@ int unregister_ftrace_function(struct ftrace_ops *ops) } EXPORT_SYMBOL_GPL(unregister_ftrace_function); +static bool is_permanent_ops_registered(void) +{ + struct ftrace_ops *op; + + do_for_each_ftrace_op(op, ftrace_ops_list) { + if (op->flags & FTRACE_OPS_FL_PERMANENT) + return true; + } while_for_each_ftrace_op(op); + + return false; +} + int ftrace_enable_sysctl(struct ctl_table *table, int write, void __user *buffer, size_t *lenp, @@ -6771,8 +6785,6 @@ ftrace_enable_sysctl(struct ctl_table *table, int write, if (ret || !write || (last_ftrace_enabled == !!ftrace_enabled)) goto out; - last_ftrace_enabled = !!ftrace_enabled; - if (ftrace_enabled) { /* we are starting ftrace again */ @@ -6783,12 +6795,19 @@ ftrace_enable_sysctl(struct ctl_table *table, int write, ftrace_startup_sysctl(); } else { + if (is_permanent_ops_registered()) { + ftrace_enabled = true; + ret = -EBUSY; + goto out; + } + /* stopping ftrace calls (just send to ftrace_stub) */ ftrace_trace_function = ftrace_stub; ftrace_shutdown_sysctl(); } + last_ftrace_enabled = !!ftrace_enabled; out: mutex_unlock(&ftrace_lock); return ret; -- cgit From 714641c3670cdc75371a7ff5bdfd5e9a170c7ffd Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Fri, 8 Nov 2019 12:25:46 -0500 Subject: ftrace: Separate out the copying of a ftrace_hash from __ftrace_hash_move() Most of the functionality of __ftrace_hash_move() can be reused, but not all of it. That is, __ftrace_hash_move() is used to simply make a new hash from an existing one, using the same size as the original. Creating a dup_hash(), where we can specify a new size will be useful when we want to create a hash with a default size, or simply copy the old one. Signed-off-by: Steven Rostedt (VMWare) --- kernel/trace/ftrace.c | 29 +++++++++++++++++------------ 1 file changed, 17 insertions(+), 12 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 89e9128652ef..76e5de8c7822 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1372,23 +1372,15 @@ ftrace_hash_rec_enable_modify(struct ftrace_ops *ops, int filter_hash); static int ftrace_hash_ipmodify_update(struct ftrace_ops *ops, struct ftrace_hash *new_hash); -static struct ftrace_hash * -__ftrace_hash_move(struct ftrace_hash *src) +static struct ftrace_hash *dup_hash(struct ftrace_hash *src, int size) { struct ftrace_func_entry *entry; - struct hlist_node *tn; - struct hlist_head *hhd; struct ftrace_hash *new_hash; - int size = src->count; + struct hlist_head *hhd; + struct hlist_node *tn; int bits = 0; int i; - /* - * If the new source is empty, just return the empty_hash. - */ - if (ftrace_hash_empty(src)) - return EMPTY_HASH; - /* * Make the hash size about 1/2 the # found */ @@ -1413,10 +1405,23 @@ __ftrace_hash_move(struct ftrace_hash *src) __add_hash_entry(new_hash, entry); } } - return new_hash; } +static struct ftrace_hash * +__ftrace_hash_move(struct ftrace_hash *src) +{ + int size = src->count; + + /* + * If the new source is empty, just return the empty_hash. + */ + if (ftrace_hash_empty(src)) + return EMPTY_HASH; + + return dup_hash(src, size); +} + static int ftrace_hash_move(struct ftrace_ops *ops, int enable, struct ftrace_hash **dst, struct ftrace_hash *src) -- cgit From 7e16f581a81759bafea04d049134b32d1a881226 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Fri, 8 Nov 2019 12:26:46 -0500 Subject: ftrace: Separate out functionality from ftrace_location_range() Create a new function called lookup_rec() from the functionality of ftrace_location_range(). The difference between lookup_rec() is that it returns the record that it finds, where as ftrace_location_range() returns only if it found a match or not. The lookup_rec() is static, and can be used for new functionality where ftrace needs to find a record of a specific address. Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/ftrace.c | 38 +++++++++++++++++++++++--------------- 1 file changed, 23 insertions(+), 15 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 76e5de8c7822..b0e7f03919de 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1541,6 +1541,26 @@ static int ftrace_cmp_recs(const void *a, const void *b) return 0; } +static struct dyn_ftrace *lookup_rec(unsigned long start, unsigned long end) +{ + struct ftrace_page *pg; + struct dyn_ftrace *rec = NULL; + struct dyn_ftrace key; + + key.ip = start; + key.flags = end; /* overload flags, as it is unsigned long */ + + for (pg = ftrace_pages_start; pg; pg = pg->next) { + if (end < pg->records[0].ip || + start >= (pg->records[pg->index - 1].ip + MCOUNT_INSN_SIZE)) + continue; + rec = bsearch(&key, pg->records, pg->index, + sizeof(struct dyn_ftrace), + ftrace_cmp_recs); + } + return rec; +} + /** * ftrace_location_range - return the first address of a traced location * if it touches the given ip range @@ -1555,23 +1575,11 @@ static int ftrace_cmp_recs(const void *a, const void *b) */ unsigned long ftrace_location_range(unsigned long start, unsigned long end) { - struct ftrace_page *pg; struct dyn_ftrace *rec; - struct dyn_ftrace key; - key.ip = start; - key.flags = end; /* overload flags, as it is unsigned long */ - - for (pg = ftrace_pages_start; pg; pg = pg->next) { - if (end < pg->records[0].ip || - start >= (pg->records[pg->index - 1].ip + MCOUNT_INSN_SIZE)) - continue; - rec = bsearch(&key, pg->records, pg->index, - sizeof(struct dyn_ftrace), - ftrace_cmp_recs); - if (rec) - return rec->ip; - } + rec = lookup_rec(start, end); + if (rec) + return rec->ip; return 0; } -- cgit From 763e34e74bb7d5c316015e2e39fcc8520bfd071c Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Fri, 8 Nov 2019 13:07:06 -0500 Subject: ftrace: Add register_ftrace_direct() Add the start of the functionality to allow other trampolines to use the ftrace mcount/fentry/nop location. This adds two new functions: register_ftrace_direct() and unregister_ftrace_direct() Both take two parameters: the first is the instruction address of where the mcount/fentry/nop exists, and the second is the trampoline to have that location called. This will handle cases where ftrace is already used on that same location, and will make it still work, where the registered direct called trampoline will get called after all the registered ftrace callers are handled. Currently, it will not allow for IP_MODIFY functions to be called at the same locations, which include some kprobes and live kernel patching. At this point, no architecture supports this. This is only the start of implementing the framework. Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/Kconfig | 8 ++ kernel/trace/ftrace.c | 269 +++++++++++++++++++++++++++++++++++++++++++++++++- 2 files changed, 272 insertions(+), 5 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index e08527f50d2a..624a05e99b0b 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -33,6 +33,9 @@ config HAVE_DYNAMIC_FTRACE config HAVE_DYNAMIC_FTRACE_WITH_REGS bool +config HAVE_DYNAMIC_FTRACE_WITH_DIRECT_CALLS + bool + config HAVE_FTRACE_MCOUNT_RECORD bool help @@ -557,6 +560,11 @@ config DYNAMIC_FTRACE_WITH_REGS depends on DYNAMIC_FTRACE depends on HAVE_DYNAMIC_FTRACE_WITH_REGS +config DYNAMIC_FTRACE_WITH_DIRECT_CALLS + def_bool y + depends on DYNAMIC_FTRACE + depends on HAVE_DYNAMIC_FTRACE_WITH_DIRECT_CALLS + config FUNCTION_PROFILER bool "Kernel function profiler" depends on FUNCTION_TRACER diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index b0e7f03919de..329a3f3789a1 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1023,6 +1023,7 @@ static bool update_all_ops; struct ftrace_func_entry { struct hlist_node hlist; unsigned long ip; + unsigned long direct; /* for direct lookup only */ }; struct ftrace_func_probe { @@ -1730,6 +1731,9 @@ static bool __ftrace_hash_rec_update(struct ftrace_ops *ops, if (FTRACE_WARN_ON(ftrace_rec_count(rec) == FTRACE_REF_MAX)) return false; + if (ops->flags & FTRACE_OPS_FL_DIRECT) + rec->flags |= FTRACE_FL_DIRECT; + /* * If there's only a single callback registered to a * function, and the ops has a trampoline registered @@ -1757,6 +1761,15 @@ static bool __ftrace_hash_rec_update(struct ftrace_ops *ops, return false; rec->flags--; + /* + * Only the internal direct_ops should have the + * DIRECT flag set. Thus, if it is removing a + * function, then that function should no longer + * be direct. + */ + if (ops->flags & FTRACE_OPS_FL_DIRECT) + rec->flags &= ~FTRACE_FL_DIRECT; + /* * If the rec had REGS enabled and the ops that is * being removed had REGS set, then see if there is @@ -2092,15 +2105,34 @@ static int ftrace_check_record(struct dyn_ftrace *rec, bool enable, bool update) * If enabling and the REGS flag does not match the REGS_EN, or * the TRAMP flag doesn't match the TRAMP_EN, then do not ignore * this record. Set flags to fail the compare against ENABLED. + * Same for direct calls. */ if (flag) { - if (!(rec->flags & FTRACE_FL_REGS) != + if (!(rec->flags & FTRACE_FL_REGS) != !(rec->flags & FTRACE_FL_REGS_EN)) flag |= FTRACE_FL_REGS; - if (!(rec->flags & FTRACE_FL_TRAMP) != + if (!(rec->flags & FTRACE_FL_TRAMP) != !(rec->flags & FTRACE_FL_TRAMP_EN)) flag |= FTRACE_FL_TRAMP; + + /* + * Direct calls are special, as count matters. + * We must test the record for direct, if the + * DIRECT and DIRECT_EN do not match, but only + * if the count is 1. That's because, if the + * count is something other than one, we do not + * want the direct enabled (it will be done via the + * direct helper). But if DIRECT_EN is set, and + * the count is not one, we need to clear it. + */ + if (ftrace_rec_count(rec) == 1) { + if (!(rec->flags & FTRACE_FL_DIRECT) != + !(rec->flags & FTRACE_FL_DIRECT_EN)) + flag |= FTRACE_FL_DIRECT; + } else if (rec->flags & FTRACE_FL_DIRECT_EN) { + flag |= FTRACE_FL_DIRECT; + } } /* If the state of this record hasn't changed, then do nothing */ @@ -2125,6 +2157,25 @@ static int ftrace_check_record(struct dyn_ftrace *rec, bool enable, bool update) else rec->flags &= ~FTRACE_FL_TRAMP_EN; } + if (flag & FTRACE_FL_DIRECT) { + /* + * If there's only one user (direct_ops helper) + * then we can call the direct function + * directly (no ftrace trampoline). + */ + if (ftrace_rec_count(rec) == 1) { + if (rec->flags & FTRACE_FL_DIRECT) + rec->flags |= FTRACE_FL_DIRECT_EN; + else + rec->flags &= ~FTRACE_FL_DIRECT_EN; + } else { + /* + * Can only call directly if there's + * only one callback to the function. + */ + rec->flags &= ~FTRACE_FL_DIRECT_EN; + } + } } /* @@ -2154,7 +2205,7 @@ static int ftrace_check_record(struct dyn_ftrace *rec, bool enable, bool update) * and REGS states. The _EN flags must be disabled though. */ rec->flags &= ~(FTRACE_FL_ENABLED | FTRACE_FL_TRAMP_EN | - FTRACE_FL_REGS_EN); + FTRACE_FL_REGS_EN | FTRACE_FL_DIRECT_EN); } ftrace_bug_type = FTRACE_BUG_NOP; @@ -2309,6 +2360,51 @@ ftrace_find_tramp_ops_new(struct dyn_ftrace *rec) return NULL; } +#ifdef CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS +/* Protected by rcu_tasks for reading, and direct_mutex for writing */ +static struct ftrace_hash *direct_functions = EMPTY_HASH; +static DEFINE_MUTEX(direct_mutex); + +/* + * Search the direct_functions hash to see if the given instruction pointer + * has a direct caller attached to it. + */ +static unsigned long find_rec_direct(unsigned long ip) +{ + struct ftrace_func_entry *entry; + + entry = __ftrace_lookup_ip(direct_functions, ip); + if (!entry) + return 0; + + return entry->direct; +} + +static void call_direct_funcs(unsigned long ip, unsigned long pip, + struct ftrace_ops *ops, struct pt_regs *regs) +{ + unsigned long addr; + + addr = find_rec_direct(ip); + if (!addr) + return; + + arch_ftrace_set_direct_caller(regs, addr); +} + +struct ftrace_ops direct_ops = { + .func = call_direct_funcs, + .flags = FTRACE_OPS_FL_IPMODIFY | FTRACE_OPS_FL_RECURSION_SAFE + | FTRACE_OPS_FL_DIRECT | FTRACE_OPS_FL_SAVE_REGS + | FTRACE_OPS_FL_PERMANENT, +}; +#else +static inline unsigned long find_rec_direct(unsigned long ip) +{ + return 0; +} +#endif /* CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS */ + /** * ftrace_get_addr_new - Get the call address to set to * @rec: The ftrace record descriptor @@ -2322,6 +2418,15 @@ ftrace_find_tramp_ops_new(struct dyn_ftrace *rec) unsigned long ftrace_get_addr_new(struct dyn_ftrace *rec) { struct ftrace_ops *ops; + unsigned long addr; + + if ((rec->flags & FTRACE_FL_DIRECT) && + (ftrace_rec_count(rec) == 1)) { + addr = find_rec_direct(rec->ip); + if (addr) + return addr; + WARN_ON_ONCE(1); + } /* Trampolines take precedence over regs */ if (rec->flags & FTRACE_FL_TRAMP) { @@ -2354,6 +2459,15 @@ unsigned long ftrace_get_addr_new(struct dyn_ftrace *rec) unsigned long ftrace_get_addr_curr(struct dyn_ftrace *rec) { struct ftrace_ops *ops; + unsigned long addr; + + /* Direct calls take precedence over trampolines */ + if (rec->flags & FTRACE_FL_DIRECT_EN) { + addr = find_rec_direct(rec->ip); + if (addr) + return addr; + WARN_ON_ONCE(1); + } /* Trampolines take precedence over regs */ if (rec->flags & FTRACE_FL_TRAMP_EN) { @@ -3465,10 +3579,11 @@ static int t_show(struct seq_file *m, void *v) if (iter->flags & FTRACE_ITER_ENABLED) { struct ftrace_ops *ops; - seq_printf(m, " (%ld)%s%s", + seq_printf(m, " (%ld)%s%s%s", ftrace_rec_count(rec), rec->flags & FTRACE_FL_REGS ? " R" : " ", - rec->flags & FTRACE_FL_IPMODIFY ? " I" : " "); + rec->flags & FTRACE_FL_IPMODIFY ? " I" : " ", + rec->flags & FTRACE_FL_DIRECT ? " D" : " "); if (rec->flags & FTRACE_FL_TRAMP_EN) { ops = ftrace_find_tramp_ops_any(rec); if (ops) { @@ -3484,6 +3599,13 @@ static int t_show(struct seq_file *m, void *v) } else { add_trampoline_func(m, NULL, rec); } + if (rec->flags & FTRACE_FL_DIRECT) { + unsigned long direct; + + direct = find_rec_direct(rec->ip); + if (direct) + seq_printf(m, "\n\tdirect-->%pS", (void *)direct); + } } seq_putc(m, '\n'); @@ -4815,6 +4937,143 @@ ftrace_set_addr(struct ftrace_ops *ops, unsigned long ip, int remove, return ftrace_set_hash(ops, NULL, 0, ip, remove, reset, enable); } +#ifdef CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS +/** + * register_ftrace_direct - Call a custom trampoline directly + * @ip: The address of the nop at the beginning of a function + * @addr: The address of the trampoline to call at @ip + * + * This is used to connect a direct call from the nop location (@ip) + * at the start of ftrace traced functions. The location that it calls + * (@addr) must be able to handle a direct call, and save the parameters + * of the function being traced, and restore them (or inject new ones + * if needed), before returning. + * + * Returns: + * 0 on success + * -EBUSY - Another direct function is already attached (there can be only one) + * -ENODEV - @ip does not point to a ftrace nop location (or not supported) + * -ENOMEM - There was an allocation failure. + */ +int register_ftrace_direct(unsigned long ip, unsigned long addr) +{ + struct ftrace_func_entry *entry; + struct ftrace_hash *free_hash = NULL; + struct dyn_ftrace *rec; + int ret = -EBUSY; + + mutex_lock(&direct_mutex); + + /* See if there's a direct function at @ip already */ + if (find_rec_direct(ip)) + goto out_unlock; + + ret = -ENODEV; + rec = lookup_rec(ip, ip); + if (!rec) + goto out_unlock; + + /* + * Check if the rec says it has a direct call but we didn't + * find one earlier? + */ + if (WARN_ON(rec->flags & FTRACE_FL_DIRECT)) + goto out_unlock; + + /* Make sure the ip points to the exact record */ + ip = rec->ip; + + ret = -ENOMEM; + if (ftrace_hash_empty(direct_functions) || + direct_functions->count > 2 * (1 << direct_functions->size_bits)) { + struct ftrace_hash *new_hash; + int size = ftrace_hash_empty(direct_functions) ? 0 : + direct_functions->count + 1; + + if (size < 32) + size = 32; + + new_hash = dup_hash(direct_functions, size); + if (!new_hash) + goto out_unlock; + + free_hash = direct_functions; + direct_functions = new_hash; + } + + entry = kmalloc(sizeof(*entry), GFP_KERNEL); + if (!entry) + goto out_unlock; + + entry->ip = ip; + entry->direct = addr; + __add_hash_entry(direct_functions, entry); + + ret = ftrace_set_filter_ip(&direct_ops, ip, 0, 0); + if (ret) + remove_hash_entry(direct_functions, entry); + + if (!ret && !(direct_ops.flags & FTRACE_OPS_FL_ENABLED)) { + ret = register_ftrace_function(&direct_ops); + if (ret) + ftrace_set_filter_ip(&direct_ops, ip, 1, 0); + } + + if (ret) + kfree(entry); + out_unlock: + mutex_unlock(&direct_mutex); + + if (free_hash) { + synchronize_rcu_tasks(); + free_ftrace_hash(free_hash); + } + + return ret; +} +EXPORT_SYMBOL_GPL(register_ftrace_direct); + +int unregister_ftrace_direct(unsigned long ip, unsigned long addr) +{ + struct ftrace_func_entry *entry; + struct dyn_ftrace *rec; + int ret = -ENODEV; + + mutex_lock(&direct_mutex); + + entry = __ftrace_lookup_ip(direct_functions, ip); + if (!entry) { + /* OK if it is off by a little */ + rec = lookup_rec(ip, ip); + if (!rec || rec->ip == ip) + goto out_unlock; + + entry = __ftrace_lookup_ip(direct_functions, rec->ip); + if (!entry) { + WARN_ON(rec->flags & FTRACE_FL_DIRECT); + goto out_unlock; + } + + WARN_ON(!(rec->flags & FTRACE_FL_DIRECT)); + } + + if (direct_functions->count == 1) + unregister_ftrace_function(&direct_ops); + + ret = ftrace_set_filter_ip(&direct_ops, ip, 1, 0); + + WARN_ON(ret); + + remove_hash_entry(direct_functions, entry); + + out_unlock: + mutex_unlock(&direct_mutex); + + return ret; +} +EXPORT_SYMBOL_GPL(unregister_ftrace_direct); +#endif /* CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS */ + /** * ftrace_set_filter_ip - set a function to filter on in ftrace by address * @ops - the ops to set the filter with -- cgit From 013bf0da0474816f57739daa006c8564ad7396a3 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Fri, 8 Nov 2019 13:11:27 -0500 Subject: ftrace: Add ftrace_find_direct_func() As function_graph tracer modifies the return address to insert a trampoline to trace the return of a function, it must be aware of a direct caller, as when it gets called, the function's return address may not be at on the stack where it expects. It may have to see if that return address points to the a direct caller and adjust if it is. Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/ftrace.c | 79 ++++++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 78 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 329a3f3789a1..c4446eabacbe 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -4938,6 +4938,46 @@ ftrace_set_addr(struct ftrace_ops *ops, unsigned long ip, int remove, } #ifdef CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS + +struct ftrace_direct_func { + struct list_head next; + unsigned long addr; + int count; +}; + +static LIST_HEAD(ftrace_direct_funcs); + +/** + * ftrace_find_direct_func - test an address if it is a registered direct caller + * @addr: The address of a registered direct caller + * + * This searches to see if a ftrace direct caller has been registered + * at a specific address, and if so, it returns a descriptor for it. + * + * This can be used by architecture code to see if an address is + * a direct caller (trampoline) attached to a fentry/mcount location. + * This is useful for the function_graph tracer, as it may need to + * do adjustments if it traced a location that also has a direct + * trampoline attached to it. + */ +struct ftrace_direct_func *ftrace_find_direct_func(unsigned long addr) +{ + struct ftrace_direct_func *entry; + bool found = false; + + /* May be called by fgraph trampoline (protected by rcu tasks) */ + list_for_each_entry_rcu(entry, &ftrace_direct_funcs, next) { + if (entry->addr == addr) { + found = true; + break; + } + } + if (found) + return entry; + + return NULL; +} + /** * register_ftrace_direct - Call a custom trampoline directly * @ip: The address of the nop at the beginning of a function @@ -4957,6 +4997,7 @@ ftrace_set_addr(struct ftrace_ops *ops, unsigned long ip, int remove, */ int register_ftrace_direct(unsigned long ip, unsigned long addr) { + struct ftrace_direct_func *direct; struct ftrace_func_entry *entry; struct ftrace_hash *free_hash = NULL; struct dyn_ftrace *rec; @@ -5005,6 +5046,18 @@ int register_ftrace_direct(unsigned long ip, unsigned long addr) if (!entry) goto out_unlock; + direct = ftrace_find_direct_func(addr); + if (!direct) { + direct = kmalloc(sizeof(*direct), GFP_KERNEL); + if (!direct) { + kfree(entry); + goto out_unlock; + } + direct->addr = addr; + direct->count = 0; + list_add_rcu(&direct->next, &ftrace_direct_funcs); + } + entry->ip = ip; entry->direct = addr; __add_hash_entry(direct_functions, entry); @@ -5019,8 +5072,20 @@ int register_ftrace_direct(unsigned long ip, unsigned long addr) ftrace_set_filter_ip(&direct_ops, ip, 1, 0); } - if (ret) + if (ret) { kfree(entry); + if (!direct->count) { + list_del_rcu(&direct->next); + synchronize_rcu_tasks(); + kfree(direct); + if (free_hash) + free_ftrace_hash(free_hash); + free_hash = NULL; + } + } else { + if (!direct->count) + direct->count++; + } out_unlock: mutex_unlock(&direct_mutex); @@ -5036,6 +5101,7 @@ EXPORT_SYMBOL_GPL(register_ftrace_direct); int unregister_ftrace_direct(unsigned long ip, unsigned long addr) { struct ftrace_func_entry *entry; + struct ftrace_direct_func *direct; struct dyn_ftrace *rec; int ret = -ENODEV; @@ -5066,6 +5132,17 @@ int unregister_ftrace_direct(unsigned long ip, unsigned long addr) remove_hash_entry(direct_functions, entry); + direct = ftrace_find_direct_func(addr); + if (!WARN_ON(!direct)) { + /* This is the good path (see the ! before WARN) */ + direct->count--; + WARN_ON(direct->count < 0); + if (!direct->count) { + list_del_rcu(&direct->next); + synchronize_rcu_tasks(); + kfree(direct); + } + } out_unlock: mutex_unlock(&direct_mutex); -- cgit From a3ad1a7e39689005cb04a4f2adb82f9d55b4724f Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Fri, 8 Nov 2019 13:12:57 -0500 Subject: ftrace/x86: Add a counter to test function_graph with direct As testing for direct calls from the function graph tracer adds a little overhead (which is a lot when tracing every function), add a counter that can be used to test if function_graph tracer needs to test for a direct caller or not. It would have been nicer if we could use a static branch, but the static branch logic fails when used within the function graph tracer trampoline. Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/ftrace.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'kernel') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index c4446eabacbe..f9456346ec66 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -2364,6 +2364,7 @@ ftrace_find_tramp_ops_new(struct dyn_ftrace *rec) /* Protected by rcu_tasks for reading, and direct_mutex for writing */ static struct ftrace_hash *direct_functions = EMPTY_HASH; static DEFINE_MUTEX(direct_mutex); +int ftrace_direct_func_count; /* * Search the direct_functions hash to see if the given instruction pointer @@ -5056,6 +5057,7 @@ int register_ftrace_direct(unsigned long ip, unsigned long addr) direct->addr = addr; direct->count = 0; list_add_rcu(&direct->next, &ftrace_direct_funcs); + ftrace_direct_func_count++; } entry->ip = ip; @@ -5081,6 +5083,7 @@ int register_ftrace_direct(unsigned long ip, unsigned long addr) if (free_hash) free_ftrace_hash(free_hash); free_hash = NULL; + ftrace_direct_func_count--; } } else { if (!direct->count) @@ -5141,6 +5144,7 @@ int unregister_ftrace_direct(unsigned long ip, unsigned long addr) list_del_rcu(&direct->next); synchronize_rcu_tasks(); kfree(direct); + ftrace_direct_func_count--; } } out_unlock: -- cgit From da537f0aef1372c5204356a7df06be8769467b7b Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Tue, 1 Oct 2019 14:38:07 -0400 Subject: ftrace: Add information on number of page groups allocated Looking for ways to shrink the size of the dyn_ftrace structure, knowing the information about how many pages and the number of groups of those pages, is useful in working out the best ways to save on memory. This adds one info print on how many groups of pages were used to allocate the ftrace dyn_ftrace structures, and also shows the number of pages and groups in the dyn_ftrace_total_info (which is used for debugging). Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/ftrace.c | 14 ++++++++++++++ kernel/trace/trace.c | 21 +++++++++++++++------ kernel/trace/trace.h | 2 ++ 3 files changed, 31 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index f9456346ec66..d2d488c43a6a 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -2991,6 +2991,8 @@ static void ftrace_shutdown_sysctl(void) static u64 ftrace_update_time; unsigned long ftrace_update_tot_cnt; +unsigned long ftrace_number_of_pages; +unsigned long ftrace_number_of_groups; static inline int ops_traces_mod(struct ftrace_ops *ops) { @@ -3115,6 +3117,9 @@ static int ftrace_allocate_records(struct ftrace_page *pg, int count) goto again; } + ftrace_number_of_pages += 1 << order; + ftrace_number_of_groups++; + cnt = (PAGE_SIZE << order) / ENTRY_SIZE; pg->size = cnt; @@ -3170,6 +3175,8 @@ ftrace_allocate_pages(unsigned long num_to_init) start_pg = pg->next; kfree(pg); pg = start_pg; + ftrace_number_of_pages -= 1 << order; + ftrace_number_of_groups--; } pr_info("ftrace: FAILED to allocate memory for functions\n"); return NULL; @@ -6173,6 +6180,8 @@ void ftrace_release_mod(struct module *mod) free_pages((unsigned long)pg->records, order); tmp_page = pg->next; kfree(pg); + ftrace_number_of_pages -= 1 << order; + ftrace_number_of_groups--; } } @@ -6514,6 +6523,8 @@ void ftrace_free_mem(struct module *mod, void *start_ptr, void *end_ptr) *last_pg = pg->next; order = get_count_order(pg->size / ENTRIES_PER_PAGE); free_pages((unsigned long)pg->records, order); + ftrace_number_of_pages -= 1 << order; + ftrace_number_of_groups--; kfree(pg); pg = container_of(last_pg, struct ftrace_page, next); if (!(*last_pg)) @@ -6569,6 +6580,9 @@ void __init ftrace_init(void) __start_mcount_loc, __stop_mcount_loc); + pr_info("ftrace: allocated %ld pages with %ld groups\n", + ftrace_number_of_pages, ftrace_number_of_groups); + set_ftrace_early_filters(); return; diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 6a0ee9178365..5ea8c7c0f2d7 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -7583,14 +7583,23 @@ static ssize_t tracing_read_dyn_info(struct file *filp, char __user *ubuf, size_t cnt, loff_t *ppos) { - unsigned long *p = filp->private_data; - char buf[64]; /* Not too big for a shallow stack */ + ssize_t ret; + char *buf; int r; - r = scnprintf(buf, 63, "%ld", *p); - buf[r++] = '\n'; + /* 256 should be plenty to hold the amount needed */ + buf = kmalloc(256, GFP_KERNEL); + if (!buf) + return -ENOMEM; - return simple_read_from_buffer(ubuf, cnt, ppos, buf, r); + r = scnprintf(buf, 256, "%ld pages:%ld groups: %ld\n", + ftrace_update_tot_cnt, + ftrace_number_of_pages, + ftrace_number_of_groups); + + ret = simple_read_from_buffer(ubuf, cnt, ppos, buf, r); + kfree(buf); + return ret; } static const struct file_operations tracing_dyn_info_fops = { @@ -8782,7 +8791,7 @@ static __init int tracer_init_tracefs(void) #ifdef CONFIG_DYNAMIC_FTRACE trace_create_file("dyn_ftrace_total_info", 0444, d_tracer, - &ftrace_update_tot_cnt, &tracing_dyn_info_fops); + NULL, &tracing_dyn_info_fops); #endif create_trace_instances(d_tracer); diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index d685c61085c0..8b590f10bc72 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -804,6 +804,8 @@ extern void trace_event_follow_fork(struct trace_array *tr, bool enable); #ifdef CONFIG_DYNAMIC_FTRACE extern unsigned long ftrace_update_tot_cnt; +extern unsigned long ftrace_number_of_pages; +extern unsigned long ftrace_number_of_groups; void ftrace_init_trace_array(struct trace_array *tr); #else static inline void ftrace_init_trace_array(struct trace_array *tr) { } -- cgit From 91edde2e6ae1dd5e33812f076f3fe4cb7ccbfdd0 Mon Sep 17 00:00:00 2001 From: "Viktor Rosendahl (BMW)" Date: Wed, 9 Oct 2019 00:08:21 +0200 Subject: ftrace: Implement fs notification for tracing_max_latency This patch implements the feature that the tracing_max_latency file, e.g. /sys/kernel/debug/tracing/tracing_max_latency will receive notifications through the fsnotify framework when a new latency is available. One particularly interesting use of this facility is when enabling threshold tracing, through /sys/kernel/debug/tracing/tracing_thresh, together with the preempt/irqsoff tracers. This makes it possible to implement a user space program that can, with equal probability, obtain traces of latencies that occur immediately after each other in spite of the fact that the preempt/irqsoff tracers operate in overwrite mode. This facility works with the hwlat, preempt/irqsoff, and wakeup tracers. The tracers may call the latency_fsnotify() from places such as __schedule() or do_idle(); this makes it impossible to call queue_work() directly without risking a deadlock. The same would happen with a softirq, kernel thread or tasklet. For this reason we use the irq_work mechanism to call queue_work(). This patch creates a new workqueue. The reason for doing this is that I wanted to use the WQ_UNBOUND and WQ_HIGHPRI flags. My thinking was that WQ_UNBOUND might help with the latency in some important cases. If we use: queue_work(system_highpri_wq, &tr->fsnotify_work); then the work will (almost) always execute on the same CPU but if we are unlucky that CPU could be too busy while there could be another CPU in the system that would be able to process the work soon enough. queue_work_on() could be used to queue the work on another CPU but it seems difficult to select the right CPU. Link: http://lkml.kernel.org/r/20191008220824.7911-2-viktor.rosendahl@gmail.com Reviewed-by: Joel Fernandes (Google) Signed-off-by: Viktor Rosendahl (BMW) [ Added max() to have one compare for max latency ] Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace.c | 75 ++++++++++++++++++++++++++++++++++++++++++++-- kernel/trace/trace.h | 18 +++++++++++ kernel/trace/trace_hwlat.c | 11 ++++--- 3 files changed, 98 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 5ea8c7c0f2d7..f093a433cb42 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -45,6 +45,9 @@ #include #include #include +#include +#include +#include #include "trace.h" #include "trace_output.h" @@ -1497,6 +1500,74 @@ static ssize_t trace_seq_to_buffer(struct trace_seq *s, void *buf, size_t cnt) } unsigned long __read_mostly tracing_thresh; +static const struct file_operations tracing_max_lat_fops; + +#if (defined(CONFIG_TRACER_MAX_TRACE) || defined(CONFIG_HWLAT_TRACER)) && \ + defined(CONFIG_FSNOTIFY) + +static struct workqueue_struct *fsnotify_wq; + +static void latency_fsnotify_workfn(struct work_struct *work) +{ + struct trace_array *tr = container_of(work, struct trace_array, + fsnotify_work); + fsnotify(tr->d_max_latency->d_inode, FS_MODIFY, + tr->d_max_latency->d_inode, FSNOTIFY_EVENT_INODE, NULL, 0); +} + +static void latency_fsnotify_workfn_irq(struct irq_work *iwork) +{ + struct trace_array *tr = container_of(iwork, struct trace_array, + fsnotify_irqwork); + queue_work(fsnotify_wq, &tr->fsnotify_work); +} + +static void trace_create_maxlat_file(struct trace_array *tr, + struct dentry *d_tracer) +{ + INIT_WORK(&tr->fsnotify_work, latency_fsnotify_workfn); + init_irq_work(&tr->fsnotify_irqwork, latency_fsnotify_workfn_irq); + tr->d_max_latency = trace_create_file("tracing_max_latency", 0644, + d_tracer, &tr->max_latency, + &tracing_max_lat_fops); +} + +__init static int latency_fsnotify_init(void) +{ + fsnotify_wq = alloc_workqueue("tr_max_lat_wq", + WQ_UNBOUND | WQ_HIGHPRI, 0); + if (!fsnotify_wq) { + pr_err("Unable to allocate tr_max_lat_wq\n"); + return -ENOMEM; + } + return 0; +} + +late_initcall_sync(latency_fsnotify_init); + +void latency_fsnotify(struct trace_array *tr) +{ + if (!fsnotify_wq) + return; + /* + * We cannot call queue_work(&tr->fsnotify_work) from here because it's + * possible that we are called from __schedule() or do_idle(), which + * could cause a deadlock. + */ + irq_work_queue(&tr->fsnotify_irqwork); +} + +/* + * (defined(CONFIG_TRACER_MAX_TRACE) || defined(CONFIG_HWLAT_TRACER)) && \ + * defined(CONFIG_FSNOTIFY) + */ +#else + +#define trace_create_maxlat_file(tr, d_tracer) \ + trace_create_file("tracing_max_latency", 0644, d_tracer, \ + &tr->max_latency, &tracing_max_lat_fops) + +#endif #ifdef CONFIG_TRACER_MAX_TRACE /* @@ -1536,6 +1607,7 @@ __update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu) /* record this tasks comm */ tracing_record_cmdline(tsk); + latency_fsnotify(tr); } /** @@ -8594,8 +8666,7 @@ init_tracer_tracefs(struct trace_array *tr, struct dentry *d_tracer) create_trace_options_dir(tr); #if defined(CONFIG_TRACER_MAX_TRACE) || defined(CONFIG_HWLAT_TRACER) - trace_create_file("tracing_max_latency", 0644, d_tracer, - &tr->max_latency, &tracing_max_lat_fops); + trace_create_maxlat_file(tr, d_tracer); #endif if (ftrace_create_function_files(tr, d_tracer)) diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 8b590f10bc72..718eb998c13e 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -16,6 +16,8 @@ #include #include #include +#include +#include #ifdef CONFIG_FTRACE_SYSCALLS #include /* For NR_SYSCALLS */ @@ -264,6 +266,11 @@ struct trace_array { #endif #if defined(CONFIG_TRACER_MAX_TRACE) || defined(CONFIG_HWLAT_TRACER) unsigned long max_latency; +#ifdef CONFIG_FSNOTIFY + struct dentry *d_max_latency; + struct work_struct fsnotify_work; + struct irq_work fsnotify_irqwork; +#endif #endif struct trace_pid_list __rcu *filtered_pids; /* @@ -786,6 +793,17 @@ void update_max_tr_single(struct trace_array *tr, struct task_struct *tsk, int cpu); #endif /* CONFIG_TRACER_MAX_TRACE */ +#if (defined(CONFIG_TRACER_MAX_TRACE) || defined(CONFIG_HWLAT_TRACER)) && \ + defined(CONFIG_FSNOTIFY) + +void latency_fsnotify(struct trace_array *tr); + +#else + +static void latency_fsnotify(struct trace_array *tr) { } + +#endif + #ifdef CONFIG_STACKTRACE void __trace_stack(struct trace_array *tr, unsigned long flags, int skip, int pc); diff --git a/kernel/trace/trace_hwlat.c b/kernel/trace/trace_hwlat.c index 862f4b0139fc..63526670605a 100644 --- a/kernel/trace/trace_hwlat.c +++ b/kernel/trace/trace_hwlat.c @@ -237,6 +237,7 @@ static int get_sample(void) /* If we exceed the threshold value, we have found a hardware latency */ if (sample > thresh || outer_sample > thresh) { struct hwlat_sample s; + u64 latency; ret = 1; @@ -253,11 +254,13 @@ static int get_sample(void) s.nmi_count = nmi_count; trace_hwlat_sample(&s); + latency = max(sample, outer_sample); + /* Keep a running maximum ever recorded hardware latency */ - if (sample > tr->max_latency) - tr->max_latency = sample; - if (outer_sample > tr->max_latency) - tr->max_latency = outer_sample; + if (latency > tr->max_latency) { + tr->max_latency = latency; + latency_fsnotify(tr); + } } out: -- cgit From 793937236d1ee032d2ee5ccc27bdd280a04e766e Mon Sep 17 00:00:00 2001 From: "Viktor Rosendahl (BMW)" Date: Wed, 9 Oct 2019 00:08:22 +0200 Subject: preemptirq_delay_test: Add the burst feature and a sysfs trigger This burst feature enables the user to generate a burst of preempt/irqsoff latencies. This makes it possible to test whether we are able to detect latencies that systematically occur very close to each other. The maximum burst size is 10. We also create 10 identical test functions, so that we get 10 different backtraces; this is useful when we want to test whether we can detect all the latencies in a burst. Otherwise, there would be no easy way of differentiating between which latency in a burst was captured by the tracer. In addition, there is a sysfs trigger, so that it's not necessary to reload the module to repeat the test. The trigger will appear as /sys/kernel/preemptirq_delay_test/trigger in sysfs. Link: http://lkml.kernel.org/r/20191008220824.7911-3-viktor.rosendahl@gmail.com Reviewed-by: Joel Fernandes (Google) Signed-off-by: Viktor Rosendahl (BMW) Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/Kconfig | 6 +- kernel/trace/preemptirq_delay_test.c | 144 ++++++++++++++++++++++++++++++----- 2 files changed, 128 insertions(+), 22 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index 624a05e99b0b..d25314bc7a1c 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -760,9 +760,9 @@ config PREEMPTIRQ_DELAY_TEST configurable delay. The module busy waits for the duration of the critical section. - For example, the following invocation forces a one-time irq-disabled - critical section for 500us: - modprobe preemptirq_delay_test test_mode=irq delay=500000 + For example, the following invocation generates a burst of three + irq-disabled critical sections for 500us: + modprobe preemptirq_delay_test test_mode=irq delay=500 burst_size=3 If unsure, say N diff --git a/kernel/trace/preemptirq_delay_test.c b/kernel/trace/preemptirq_delay_test.c index d8765c952fab..31c0fad4cb9e 100644 --- a/kernel/trace/preemptirq_delay_test.c +++ b/kernel/trace/preemptirq_delay_test.c @@ -10,18 +10,25 @@ #include #include #include +#include #include #include #include #include +#include static ulong delay = 100; -static char test_mode[10] = "irq"; +static char test_mode[12] = "irq"; +static uint burst_size = 1; -module_param_named(delay, delay, ulong, S_IRUGO); -module_param_string(test_mode, test_mode, 10, S_IRUGO); -MODULE_PARM_DESC(delay, "Period in microseconds (100 uS default)"); -MODULE_PARM_DESC(test_mode, "Mode of the test such as preempt or irq (default irq)"); +module_param_named(delay, delay, ulong, 0444); +module_param_string(test_mode, test_mode, 12, 0444); +module_param_named(burst_size, burst_size, uint, 0444); +MODULE_PARM_DESC(delay, "Period in microseconds (100 us default)"); +MODULE_PARM_DESC(test_mode, "Mode of the test such as preempt, irq, or alternate (default irq)"); +MODULE_PARM_DESC(burst_size, "The size of a burst (default 1)"); + +#define MIN(x, y) ((x) < (y) ? (x) : (y)) static void busy_wait(ulong time) { @@ -34,37 +41,136 @@ static void busy_wait(ulong time) } while ((end - start) < (time * 1000)); } -static int preemptirq_delay_run(void *data) +static __always_inline void irqoff_test(void) { unsigned long flags; + local_irq_save(flags); + busy_wait(delay); + local_irq_restore(flags); +} - if (!strcmp(test_mode, "irq")) { - local_irq_save(flags); - busy_wait(delay); - local_irq_restore(flags); - } else if (!strcmp(test_mode, "preempt")) { - preempt_disable(); - busy_wait(delay); - preempt_enable(); +static __always_inline void preemptoff_test(void) +{ + preempt_disable(); + busy_wait(delay); + preempt_enable(); +} + +static void execute_preemptirqtest(int idx) +{ + if (!strcmp(test_mode, "irq")) + irqoff_test(); + else if (!strcmp(test_mode, "preempt")) + preemptoff_test(); + else if (!strcmp(test_mode, "alternate")) { + if (idx % 2 == 0) + irqoff_test(); + else + preemptoff_test(); } +} + +#define DECLARE_TESTFN(POSTFIX) \ + static void preemptirqtest_##POSTFIX(int idx) \ + { \ + execute_preemptirqtest(idx); \ + } \ +/* + * We create 10 different functions, so that we can get 10 different + * backtraces. + */ +DECLARE_TESTFN(0) +DECLARE_TESTFN(1) +DECLARE_TESTFN(2) +DECLARE_TESTFN(3) +DECLARE_TESTFN(4) +DECLARE_TESTFN(5) +DECLARE_TESTFN(6) +DECLARE_TESTFN(7) +DECLARE_TESTFN(8) +DECLARE_TESTFN(9) + +static void (*testfuncs[])(int) = { + preemptirqtest_0, + preemptirqtest_1, + preemptirqtest_2, + preemptirqtest_3, + preemptirqtest_4, + preemptirqtest_5, + preemptirqtest_6, + preemptirqtest_7, + preemptirqtest_8, + preemptirqtest_9, +}; + +#define NR_TEST_FUNCS ARRAY_SIZE(testfuncs) + +static int preemptirq_delay_run(void *data) +{ + int i; + int s = MIN(burst_size, NR_TEST_FUNCS); + + for (i = 0; i < s; i++) + (testfuncs[i])(i); return 0; } -static int __init preemptirq_delay_init(void) +static struct task_struct *preemptirq_start_test(void) { char task_name[50]; - struct task_struct *test_task; snprintf(task_name, sizeof(task_name), "%s_test", test_mode); + return kthread_run(preemptirq_delay_run, NULL, task_name); +} + + +static ssize_t trigger_store(struct kobject *kobj, struct kobj_attribute *attr, + const char *buf, size_t count) +{ + preemptirq_start_test(); + return count; +} + +static struct kobj_attribute trigger_attribute = + __ATTR(trigger, 0200, NULL, trigger_store); + +static struct attribute *attrs[] = { + &trigger_attribute.attr, + NULL, +}; + +static struct attribute_group attr_group = { + .attrs = attrs, +}; + +static struct kobject *preemptirq_delay_kobj; + +static int __init preemptirq_delay_init(void) +{ + struct task_struct *test_task; + int retval; + + test_task = preemptirq_start_test(); + retval = PTR_ERR_OR_ZERO(test_task); + if (retval != 0) + return retval; + + preemptirq_delay_kobj = kobject_create_and_add("preemptirq_delay_test", + kernel_kobj); + if (!preemptirq_delay_kobj) + return -ENOMEM; + + retval = sysfs_create_group(preemptirq_delay_kobj, &attr_group); + if (retval) + kobject_put(preemptirq_delay_kobj); - test_task = kthread_run(preemptirq_delay_run, NULL, task_name); - return PTR_ERR_OR_ZERO(test_task); + return retval; } static void __exit preemptirq_delay_exit(void) { - return; + kobject_put(preemptirq_delay_kobj); } module_init(preemptirq_delay_init) -- cgit From 9c34fc4b7e903117cc27712b9e6c8690debb7e95 Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Tue, 15 Oct 2019 21:18:20 +0200 Subject: tracing: Use CONFIG_PREEMPTION CONFIG_PREEMPTION is selected by CONFIG_PREEMPT and by CONFIG_PREEMPT_RT. Both PREEMPT and PREEMPT_RT require the same functionality which today depends on CONFIG_PREEMPT. Add additional header output for PREEMPT_RT. Link: http://lkml.kernel.org/r/20191015191821.11479-34-bigeasy@linutronix.de Cc: Ingo Molnar Signed-off-by: Sebastian Andrzej Siewior Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'kernel') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index f093a433cb42..db7d06a26861 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -3726,6 +3726,8 @@ print_trace_header(struct seq_file *m, struct trace_iterator *iter) "desktop", #elif defined(CONFIG_PREEMPT) "preempt", +#elif defined(CONFIG_PREEMPT_RT) + "preempt_rt", #else "unknown", #endif -- cgit From 6dff4d7dd3e0158688683a17dd792861aa9d61e2 Mon Sep 17 00:00:00 2001 From: Ben Dooks Date: Tue, 15 Oct 2019 13:10:12 +0100 Subject: tracing: Make internal ftrace events static The event_class_ftrace_##call and event_##call do not seem to be used outside of trace_export.c so make them both static to avoid a number of sparse warnings: kernel/trace/trace_entries.h:59:1: warning: symbol 'event_class_ftrace_function' was not declared. Should it be static? kernel/trace/trace_entries.h:59:1: warning: symbol '__event_function' was not declared. Should it be static? kernel/trace/trace_entries.h:77:1: warning: symbol 'event_class_ftrace_funcgraph_entry' was not declared. Should it be static? kernel/trace/trace_entries.h:77:1: warning: symbol '__event_funcgraph_entry' was not declared. Should it be static? kernel/trace/trace_entries.h:93:1: warning: symbol 'event_class_ftrace_funcgraph_exit' was not declared. Should it be static? kernel/trace/trace_entries.h:93:1: warning: symbol '__event_funcgraph_exit' was not declared. Should it be static? kernel/trace/trace_entries.h:129:1: warning: symbol 'event_class_ftrace_context_switch' was not declared. Should it be static? kernel/trace/trace_entries.h:129:1: warning: symbol '__event_context_switch' was not declared. Should it be static? kernel/trace/trace_entries.h:149:1: warning: symbol 'event_class_ftrace_wakeup' was not declared. Should it be static? kernel/trace/trace_entries.h:149:1: warning: symbol '__event_wakeup' was not declared. Should it be static? kernel/trace/trace_entries.h:171:1: warning: symbol 'event_class_ftrace_kernel_stack' was not declared. Should it be static? kernel/trace/trace_entries.h:171:1: warning: symbol '__event_kernel_stack' was not declared. Should it be static? kernel/trace/trace_entries.h:191:1: warning: symbol 'event_class_ftrace_user_stack' was not declared. Should it be static? kernel/trace/trace_entries.h:191:1: warning: symbol '__event_user_stack' was not declared. Should it be static? kernel/trace/trace_entries.h:214:1: warning: symbol 'event_class_ftrace_bprint' was not declared. Should it be static? kernel/trace/trace_entries.h:214:1: warning: symbol '__event_bprint' was not declared. Should it be static? kernel/trace/trace_entries.h:230:1: warning: symbol 'event_class_ftrace_print' was not declared. Should it be static? kernel/trace/trace_entries.h:230:1: warning: symbol '__event_print' was not declared. Should it be static? kernel/trace/trace_entries.h:247:1: warning: symbol 'event_class_ftrace_raw_data' was not declared. Should it be static? kernel/trace/trace_entries.h:247:1: warning: symbol '__event_raw_data' was not declared. Should it be static? kernel/trace/trace_entries.h:262:1: warning: symbol 'event_class_ftrace_bputs' was not declared. Should it be static? kernel/trace/trace_entries.h:262:1: warning: symbol '__event_bputs' was not declared. Should it be static? kernel/trace/trace_entries.h:277:1: warning: symbol 'event_class_ftrace_mmiotrace_rw' was not declared. Should it be static? kernel/trace/trace_entries.h:277:1: warning: symbol '__event_mmiotrace_rw' was not declared. Should it be static? kernel/trace/trace_entries.h:298:1: warning: symbol 'event_class_ftrace_mmiotrace_map' was not declared. Should it be static? kernel/trace/trace_entries.h:298:1: warning: symbol '__event_mmiotrace_map' was not declared. Should it be static? kernel/trace/trace_entries.h:322:1: warning: symbol 'event_class_ftrace_branch' was not declared. Should it be static? kernel/trace/trace_entries.h:322:1: warning: symbol '__event_branch' was not declared. Should it be static? kernel/trace/trace_entries.h:343:1: warning: symbol 'event_class_ftrace_hwlat' was not declared. Should it be static? kernel/trace/trace_entries.h:343:1: warning: symbol '__event_hwlat' was not declared. Should it be static? Link: http://lkml.kernel.org/r/20191015121012.18824-1-ben.dooks@codethink.co.uk Signed-off-by: Ben Dooks Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_export.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace_export.c b/kernel/trace/trace_export.c index 45630a76ed3a..2e6d2e9741cc 100644 --- a/kernel/trace/trace_export.c +++ b/kernel/trace/trace_export.c @@ -171,7 +171,7 @@ ftrace_define_fields_##name(struct trace_event_call *event_call) \ #define FTRACE_ENTRY_REG(call, struct_name, etype, tstruct, print, filter,\ regfn) \ \ -struct trace_event_class __refdata event_class_ftrace_##call = { \ +static struct trace_event_class __refdata event_class_ftrace_##call = { \ .system = __stringify(TRACE_SYSTEM), \ .define_fields = ftrace_define_fields_##call, \ .fields = LIST_HEAD_INIT(event_class_ftrace_##call.fields),\ @@ -187,7 +187,7 @@ struct trace_event_call __used event_##call = { \ .print_fmt = print, \ .flags = TRACE_EVENT_FL_IGNORE_ENABLE, \ }; \ -struct trace_event_call __used \ +static struct trace_event_call __used \ __attribute__((section("_ftrace_events"))) *__event_##call = &event_##call; #undef FTRACE_ENTRY -- cgit From 2d6425af61166e026e7476db64f70f1266127b1d Mon Sep 17 00:00:00 2001 From: Divya Indi Date: Wed, 14 Aug 2019 10:55:23 -0700 Subject: tracing: Declare newly exported APIs in include/linux/trace.h Declare the newly introduced and exported APIs in the header file - include/linux/trace.h. Moving previous declarations from kernel/trace/trace.h to include/linux/trace.h. Link: http://lkml.kernel.org/r/1565805327-579-2-git-send-email-divya.indi@oracle.com Signed-off-by: Divya Indi Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace.h | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 718eb998c13e..90cba68c8b50 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -11,6 +11,7 @@ #include #include #include +#include #include #include #include @@ -873,8 +874,6 @@ trace_vprintk(unsigned long ip, const char *fmt, va_list args); extern int trace_array_vprintk(struct trace_array *tr, unsigned long ip, const char *fmt, va_list args); -int trace_array_printk(struct trace_array *tr, - unsigned long ip, const char *fmt, ...); int trace_array_printk_buf(struct ring_buffer *buffer, unsigned long ip, const char *fmt, ...); void trace_printk_seq(struct trace_seq *s); @@ -1890,7 +1889,6 @@ extern const char *__start___tracepoint_str[]; extern const char *__stop___tracepoint_str[]; void trace_printk_control(bool enabled); -void trace_printk_init_buffers(void); void trace_printk_start_comm(void); int trace_keep_overwrite(struct tracer *tracer, u32 mask, int set); int set_tracer_flag(struct trace_array *tr, unsigned int mask, int enabled); -- cgit From e585e6469d6f476b82aa148dc44aaf7ae269a4e2 Mon Sep 17 00:00:00 2001 From: Divya Indi Date: Wed, 14 Aug 2019 10:55:24 -0700 Subject: tracing: Verify if trace array exists before destroying it. A trace array can be destroyed from userspace or kernel. Verify if the trace array exists before proceeding to destroy/remove it. Link: http://lkml.kernel.org/r/1565805327-579-3-git-send-email-divya.indi@oracle.com Reviewed-by: Aruna Ramakrishna Signed-off-by: Divya Indi [ Removed unneeded braces ] Signed-off-by: Steven Rostedt (VMware) --- kernel/module.c | 6 +++++- kernel/trace/trace.c | 15 ++++++++++++--- 2 files changed, 17 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/module.c b/kernel/module.c index ff2d7359a418..6e2fd40a6ed9 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -3728,7 +3728,6 @@ static int complete_formation(struct module *mod, struct load_info *info) module_enable_ro(mod, false); module_enable_nx(mod); - module_enable_x(mod); /* Mark state as coming so strong_try_module_get() ignores us, * but kallsyms etc. can see us. */ @@ -3751,6 +3750,11 @@ static int prepare_coming_module(struct module *mod) if (err) return err; + /* Make module executable after ftrace is enabled */ + mutex_lock(&module_mutex); + module_enable_x(mod); + mutex_unlock(&module_mutex); + blocking_notifier_call_chain(&module_notify_list, MODULE_STATE_COMING, mod); return 0; diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index db7d06a26861..fa4f742fc449 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -8556,17 +8556,26 @@ static int __remove_instance(struct trace_array *tr) return 0; } -int trace_array_destroy(struct trace_array *tr) +int trace_array_destroy(struct trace_array *this_tr) { + struct trace_array *tr; int ret; - if (!tr) + if (!this_tr) return -EINVAL; mutex_lock(&event_mutex); mutex_lock(&trace_types_lock); - ret = __remove_instance(tr); + ret = -ENODEV; + + /* Making sure trace array exists before destroying it. */ + list_for_each_entry(tr, &ftrace_trace_arrays, list) { + if (tr == this_tr) { + ret = __remove_instance(tr); + break; + } + } mutex_unlock(&trace_types_lock); mutex_unlock(&event_mutex); -- cgit From 953ae45a0c25e09428d4a03d7654f97ab8a36647 Mon Sep 17 00:00:00 2001 From: Divya Indi Date: Wed, 14 Aug 2019 10:55:25 -0700 Subject: tracing: Adding NULL checks for trace_array descriptor pointer As part of commit f45d1225adb0 ("tracing: Kernel access to Ftrace instances") we exported certain functions. Here, we are adding some additional NULL checks to ensure safe usage by users of these APIs. Link: http://lkml.kernel.org/r/1565805327-579-4-git-send-email-divya.indi@oracle.com Signed-off-by: Divya Indi Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace.c | 3 +++ kernel/trace/trace_events.c | 2 ++ 2 files changed, 5 insertions(+) (limited to 'kernel') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index fa4f742fc449..79fe4d6ecbd8 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -3297,6 +3297,9 @@ int trace_array_printk(struct trace_array *tr, if (!(global_trace.trace_flags & TRACE_ITER_PRINTK)) return 0; + if (!tr) + return -ENOENT; + va_start(ap, fmt); ret = trace_array_vprintk(tr, ip, fmt, ap); va_end(ap); diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index fba87d10f0c1..2a3ac2365445 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -793,6 +793,8 @@ int ftrace_set_clr_event(struct trace_array *tr, char *buf, int set) char *event = NULL, *sub = NULL, *match; int ret; + if (!tr) + return -ENOENT; /* * The buf format can be : * *: means any event by that name. -- cgit From b83b43ffc6e4b514ca034a0fbdee01322e2f7022 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Tue, 15 Oct 2019 09:00:55 -0400 Subject: fgraph: Fix function type mismatches of ftrace_graph_return using ftrace_stub The C compiler is allowing more checks to make sure that function pointers are assigned to the correct prototype function. Unfortunately, the function graph tracer uses a special name with its assigned ftrace_graph_return function pointer that maps to a stub function used by the function tracer (ftrace_stub). The ftrace_graph_return variable is compared to the ftrace_stub in some archs to know if the function graph tracer is enabled or not. This means we can not just simply create a new function stub that compares it without modifying all the archs. Instead, have the linker script create a function_graph_stub that maps to ftrace_stub, and this way we can define the prototype for it to match the prototype of ftrace_graph_return, and make the compiler checks all happy! Link: http://lkml.kernel.org/r/20191015090055.789a0aed@gandalf.local.home Cc: linux-sh@vger.kernel.org Cc: Yoshinori Sato Cc: Rich Felker Reported-by: Sami Tolvanen Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/fgraph.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/fgraph.c b/kernel/trace/fgraph.c index 7950a0356042..fa3ce10d0405 100644 --- a/kernel/trace/fgraph.c +++ b/kernel/trace/fgraph.c @@ -332,9 +332,14 @@ int ftrace_graph_entry_stub(struct ftrace_graph_ent *trace) return 0; } +/* + * Simply points to ftrace_stub, but with the proper protocol. + * Defined by the linker script in linux/vmlinux.lds.h + */ +extern void ftrace_graph_stub(struct ftrace_graph_ret *); + /* The callbacks that hook a function */ -trace_func_graph_ret_t ftrace_graph_return = - (trace_func_graph_ret_t)ftrace_stub; +trace_func_graph_ret_t ftrace_graph_return = ftrace_graph_stub; trace_func_graph_ent_t ftrace_graph_entry = ftrace_graph_entry_stub; static trace_func_graph_ent_t __ftrace_graph_entry = ftrace_graph_entry_stub; @@ -614,7 +619,7 @@ void unregister_ftrace_graph(struct fgraph_ops *gops) goto out; ftrace_graph_active--; - ftrace_graph_return = (trace_func_graph_ret_t)ftrace_stub; + ftrace_graph_return = ftrace_graph_stub; ftrace_graph_entry = ftrace_graph_entry_stub; __ftrace_graph_entry = ftrace_graph_entry_stub; ftrace_shutdown(&graph_ops, FTRACE_STOP_FUNC_RET); -- cgit From 80042c8f06bf5a7b87a63deaa3deb56f2cd52645 Mon Sep 17 00:00:00 2001 From: Andy Shevchenko Date: Mon, 7 Oct 2019 16:56:56 +0300 Subject: tracing: Use generic type for comparator function Comparator function type, cmp_func_t, is defined in the types.h, use it in the code. Link: http://lkml.kernel.org/r/20191007135656.37734-3-andriy.shevchenko@linux.intel.com Signed-off-by: Andy Shevchenko Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/ftrace.c | 12 ++++++------ kernel/trace/trace_branch.c | 8 ++++---- kernel/trace/trace_stat.c | 6 ++---- kernel/trace/trace_stat.h | 2 +- 4 files changed, 13 insertions(+), 15 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index d2d488c43a6a..82ef8d60a42b 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -465,10 +465,10 @@ static void *function_stat_start(struct tracer_stat *trace) #ifdef CONFIG_FUNCTION_GRAPH_TRACER /* function graph compares on total time */ -static int function_stat_cmp(void *p1, void *p2) +static int function_stat_cmp(const void *p1, const void *p2) { - struct ftrace_profile *a = p1; - struct ftrace_profile *b = p2; + const struct ftrace_profile *a = p1; + const struct ftrace_profile *b = p2; if (a->time < b->time) return -1; @@ -479,10 +479,10 @@ static int function_stat_cmp(void *p1, void *p2) } #else /* not function graph compares against hits */ -static int function_stat_cmp(void *p1, void *p2) +static int function_stat_cmp(const void *p1, const void *p2) { - struct ftrace_profile *a = p1; - struct ftrace_profile *b = p2; + const struct ftrace_profile *a = p1; + const struct ftrace_profile *b = p2; if (a->counter < b->counter) return -1; diff --git a/kernel/trace/trace_branch.c b/kernel/trace/trace_branch.c index 3ea65cdff30d..88e158d27965 100644 --- a/kernel/trace/trace_branch.c +++ b/kernel/trace/trace_branch.c @@ -244,7 +244,7 @@ static int annotated_branch_stat_headers(struct seq_file *m) return 0; } -static inline long get_incorrect_percent(struct ftrace_branch_data *p) +static inline long get_incorrect_percent(const struct ftrace_branch_data *p) { long percent; @@ -332,10 +332,10 @@ annotated_branch_stat_next(void *v, int idx) return p; } -static int annotated_branch_stat_cmp(void *p1, void *p2) +static int annotated_branch_stat_cmp(const void *p1, const void *p2) { - struct ftrace_branch_data *a = p1; - struct ftrace_branch_data *b = p2; + const struct ftrace_branch_data *a = p1; + const struct ftrace_branch_data *b = p2; long percent_a, percent_b; diff --git a/kernel/trace/trace_stat.c b/kernel/trace/trace_stat.c index 9ab0a1a7ad5e..874f1274cf99 100644 --- a/kernel/trace/trace_stat.c +++ b/kernel/trace/trace_stat.c @@ -72,9 +72,7 @@ static void destroy_session(struct stat_session *session) kfree(session); } -typedef int (*cmp_stat_t)(void *, void *); - -static int insert_stat(struct rb_root *root, void *stat, cmp_stat_t cmp) +static int insert_stat(struct rb_root *root, void *stat, cmp_func_t cmp) { struct rb_node **new = &(root->rb_node), *parent = NULL; struct stat_node *data; @@ -112,7 +110,7 @@ static int insert_stat(struct rb_root *root, void *stat, cmp_stat_t cmp) * This one will force an insertion as right-most node * in the rbtree. */ -static int dummy_cmp(void *p1, void *p2) +static int dummy_cmp(const void *p1, const void *p2) { return -1; } diff --git a/kernel/trace/trace_stat.h b/kernel/trace/trace_stat.h index 8786d17caf49..31d7dc5bf1db 100644 --- a/kernel/trace/trace_stat.h +++ b/kernel/trace/trace_stat.h @@ -16,7 +16,7 @@ struct tracer_stat { void *(*stat_start)(struct tracer_stat *trace); void *(*stat_next)(void *prev, int idx); /* Compare two entries for stats sorting */ - int (*stat_cmp)(void *p1, void *p2); + cmp_func_t stat_cmp; /* Print a stat entry */ int (*stat_show)(struct seq_file *s, void *p); /* Release an entry */ -- cgit From 0c3c86bdc691c794a6154f8515b7fa82c82dfc4d Mon Sep 17 00:00:00 2001 From: "Srivatsa S. Bhat (VMware)" Date: Thu, 10 Oct 2019 11:51:17 -0700 Subject: tracing/hwlat: Fix a few trivial nits Update the source file name in the comments, and fix a grammatical error. Link: http://lkml.kernel.org/r/157073346821.17189.8946944856026592247.stgit@srivatsa-ubuntu Signed-off-by: Srivatsa S. Bhat (VMware) Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_hwlat.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace_hwlat.c b/kernel/trace/trace_hwlat.c index 63526670605a..6638d63f0921 100644 --- a/kernel/trace/trace_hwlat.c +++ b/kernel/trace/trace_hwlat.c @@ -1,6 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 /* - * trace_hwlatdetect.c - A simple Hardware Latency detector. + * trace_hwlat.c - A simple Hardware Latency detector. * * Use this tracer to detect large system latencies induced by the behavior of * certain underlying system hardware or firmware, independent of Linux itself. @@ -279,7 +279,7 @@ static void move_to_next_cpu(void) return; /* * If for some reason the user modifies the CPU affinity - * of this thread, than stop migrating for the duration + * of this thread, then stop migrating for the duration * of the current test. */ if (!cpumask_equal(current_mask, current->cpus_ptr)) -- cgit From 6ee40511cb838f9ced002dff7131bca87e3ccbdd Mon Sep 17 00:00:00 2001 From: Yuming Han Date: Thu, 24 Oct 2019 11:34:30 +0800 Subject: tracing: use kvcalloc for tgid_map array allocation Fail to allocate memory for tgid_map, because it requires order-6 page. detail as: c3 sh: page allocation failure: order:6, mode:0x140c0c0(GFP_KERNEL), nodemask=(null) c3 sh cpuset=/ mems_allowed=0 c3 CPU: 3 PID: 5632 Comm: sh Tainted: G W O 4.14.133+ #10 c3 Hardware name: Generic DT based system c3 Backtrace: c3 [] (dump_backtrace) from [](show_stack+0x18/0x1c) c3 [] (show_stack) from [](dump_stack+0x84/0xa4) c3 [] (dump_stack) from [](warn_alloc+0xc4/0x19c) c3 [] (warn_alloc) from [](__alloc_pages_nodemask+0xd18/0xf28) c3 [] (__alloc_pages_nodemask) from [](kmalloc_order+0x20/0x38) c3 [] (kmalloc_order) from [](kmalloc_order_trace+0x24/0x108) c3 [] (kmalloc_order_trace) from [](set_tracer_flag+0xb0/0x158) c3 [] (set_tracer_flag) from [](trace_options_core_write+0x7c/0xcc) c3 [] (trace_options_core_write) from [](__vfs_write+0x40/0x14c) c3 [] (__vfs_write) from [](vfs_write+0xc4/0x198) c3 [] (vfs_write) from [](SyS_write+0x6c/0xd0) c3 [] (SyS_write) from [](ret_fast_syscall+0x0/0x54) Switch to use kvcalloc to avoid unexpected allocation failures. Link: http://lkml.kernel.org/r/1571888070-24425-1-git-send-email-chunyan.zhang@unisoc.com Signed-off-by: Yuming Han Signed-off-by: Chunyan Zhang Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 79fe4d6ecbd8..42659ce6ac0c 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -4686,7 +4686,7 @@ int set_tracer_flag(struct trace_array *tr, unsigned int mask, int enabled) if (mask == TRACE_ITER_RECORD_TGID) { if (!tgid_map) - tgid_map = kcalloc(PID_MAX_DEFAULT + 1, + tgid_map = kvcalloc(PID_MAX_DEFAULT + 1, sizeof(*tgid_map), GFP_KERNEL); if (!tgid_map) { -- cgit From c7411a1a126f649be71526a36d4afac9e5aefa13 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 29 Oct 2019 17:31:44 +0900 Subject: tracing/kprobe: Check whether the non-suffixed symbol is notrace Check whether the non-suffixed symbol is notrace, since suffixed symbols are generated by the compilers for optimization. Based on these suffixed symbols, notrace check might not work because some of them are just a partial code of the original function. (e.g. cold-cache (unlikely) code is separated from original function as FUNCTION.cold.XX) For example, without this fix, # echo p device_add.cold.67 > /sys/kernel/debug/tracing/kprobe_events sh: write error: Invalid argument # cat /sys/kernel/debug/tracing/error_log [ 135.491035] trace_kprobe: error: Failed to register probe event Command: p device_add.cold.67 ^ # dmesg | tail -n 1 [ 135.488599] trace_kprobe: Could not probe notrace function device_add.cold.67 With this, # echo p device_add.cold.66 > /sys/kernel/debug/tracing/kprobe_events # cat /sys/kernel/debug/kprobes/list ffffffff81599de9 k device_add.cold.66+0x0 [DISABLED] Actually, kprobe blacklist already did similar thing, see within_kprobe_blacklist(). Link: http://lkml.kernel.org/r/157233790394.6706.18243942030937189679.stgit@devnote2 Fixes: 45408c4f9250 ("tracing: kprobes: Prohibit probing on notrace function") Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_kprobe.c | 27 ++++++++++++++++++++++++--- 1 file changed, 24 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 1552a95c743b..7f890262c8a3 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -435,11 +435,10 @@ static int disable_trace_kprobe(struct trace_event_call *call, #if defined(CONFIG_KPROBES_ON_FTRACE) && \ !defined(CONFIG_KPROBE_EVENTS_ON_NOTRACE) -static bool within_notrace_func(struct trace_kprobe *tk) +static bool __within_notrace_func(unsigned long addr) { - unsigned long offset, size, addr; + unsigned long offset, size; - addr = trace_kprobe_address(tk); if (!addr || !kallsyms_lookup_size_offset(addr, &size, &offset)) return false; @@ -452,6 +451,28 @@ static bool within_notrace_func(struct trace_kprobe *tk) */ return !ftrace_location_range(addr, addr + size - 1); } + +static bool within_notrace_func(struct trace_kprobe *tk) +{ + unsigned long addr = addr = trace_kprobe_address(tk); + char symname[KSYM_NAME_LEN], *p; + + if (!__within_notrace_func(addr)) + return false; + + /* Check if the address is on a suffixed-symbol */ + if (!lookup_symbol_name(addr, symname)) { + p = strchr(symname, '.'); + if (!p) + return true; + *p = '\0'; + addr = (unsigned long)kprobe_lookup_name(symname, 0); + if (addr) + return __within_notrace_func(addr); + } + + return true; +} #else #define within_notrace_func(tk) (false) #endif -- cgit From ef56e047b2bd4dabb801fd073dfcab5f40de5f78 Mon Sep 17 00:00:00 2001 From: Piotr Maziarz Date: Thu, 7 Nov 2019 13:45:38 +0100 Subject: tracing: Use seq_buf_hex_dump() to dump buffers Without this, buffers can be printed with __print_array macro that has no formatting options and can be hard to read. The other way is to mimic formatting capability with multiple calls of trace event with one call per row which gives performance impact and different timestamp in each row. Link: http://lkml.kernel.org/r/1573130738-29390-2-git-send-email-piotrx.maziarz@linux.intel.com Signed-off-by: Piotr Maziarz Signed-off-by: Cezary Rojewski Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_output.c | 15 +++++++++++++++ kernel/trace/trace_seq.c | 30 ++++++++++++++++++++++++++++++ 2 files changed, 45 insertions(+) (limited to 'kernel') diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c index d54ce252b05a..d9b4b7c22db4 100644 --- a/kernel/trace/trace_output.c +++ b/kernel/trace/trace_output.c @@ -274,6 +274,21 @@ trace_print_array_seq(struct trace_seq *p, const void *buf, int count, } EXPORT_SYMBOL(trace_print_array_seq); +const char * +trace_print_hex_dump_seq(struct trace_seq *p, const char *prefix_str, + int prefix_type, int rowsize, int groupsize, + const void *buf, size_t len, bool ascii) +{ + const char *ret = trace_seq_buffer_ptr(p); + + trace_seq_putc(p, '\n'); + trace_seq_hex_dump(p, prefix_str, prefix_type, + rowsize, groupsize, buf, len, ascii); + trace_seq_putc(p, 0); + return ret; +} +EXPORT_SYMBOL(trace_print_hex_dump_seq); + int trace_raw_output_prep(struct trace_iterator *iter, struct trace_event *trace_event) { diff --git a/kernel/trace/trace_seq.c b/kernel/trace/trace_seq.c index 6b1c562ffdaf..344e4c1aa09c 100644 --- a/kernel/trace/trace_seq.c +++ b/kernel/trace/trace_seq.c @@ -376,3 +376,33 @@ int trace_seq_to_user(struct trace_seq *s, char __user *ubuf, int cnt) return seq_buf_to_user(&s->seq, ubuf, cnt); } EXPORT_SYMBOL_GPL(trace_seq_to_user); + +int trace_seq_hex_dump(struct trace_seq *s, const char *prefix_str, + int prefix_type, int rowsize, int groupsize, + const void *buf, size_t len, bool ascii) +{ + unsigned int save_len = s->seq.len; + + if (s->full) + return 0; + + __trace_seq_init(s); + + if (TRACE_SEQ_BUF_LEFT(s) < 1) { + s->full = 1; + return 0; + } + + seq_buf_hex_dump(&(s->seq), prefix_str, + prefix_type, rowsize, groupsize, + buf, len, ascii); + + if (unlikely(seq_buf_has_overflowed(&s->seq))) { + s->seq.len = save_len; + s->full = 1; + return 0; + } + + return 1; +} +EXPORT_SYMBOL(trace_seq_hex_dump); -- cgit From 9b4712044d059e7842aaeeafd7c7a7ee88c589db Mon Sep 17 00:00:00 2001 From: Borislav Petkov Date: Tue, 12 Nov 2019 18:42:19 +0100 Subject: tracing: Remove stray tab in TRACE_EVAL_MAP_FILE's help text There was a stray tab in the help text of the aforementioned config option which showed like this: The "print fmt" of the trace events will show the enum/sizeof names instead of their values. This can cause problems for user space tools ... in menuconfig. Remove it and end a sentence with a fullstop. No functional changes. Link: http://lkml.kernel.org/r/20191112174219.10933-1-bp@alien8.de Signed-off-by: Borislav Petkov Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index d25314bc7a1c..b872716bb2a0 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -771,7 +771,7 @@ config TRACE_EVAL_MAP_FILE depends on TRACING help The "print fmt" of the trace events will show the enum/sizeof names - instead of their values. This can cause problems for user space tools + instead of their values. This can cause problems for user space tools that use this string to parse the raw data as user space does not know how to convert the string to its value. @@ -792,7 +792,7 @@ config TRACE_EVAL_MAP_FILE they are needed for the "eval_map" file. Enabling this option will increase the memory footprint of the running kernel. - If unsure, say N + If unsure, say N. config GCOV_PROFILE_FTRACE bool "Enable GCOV profiling on ftrace subsystem" -- cgit From 36b3615dc3b625c8b587f34e413a600f7ac16403 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Thu, 14 Nov 2019 22:43:58 -0500 Subject: tracing: Add missing "inline" in stub function of latency_fsnotify() The latency_fsnotify() stub when the function is not defined, was missing the "inline". Link: https://lore.kernel.org/r/20191115140213.74c5efe7@canb.auug.org.au Reported-by: Stephen Rothwell Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 90cba68c8b50..2df8aed6a8f0 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -801,7 +801,7 @@ void latency_fsnotify(struct trace_array *tr); #else -static void latency_fsnotify(struct trace_array *tr) { } +static inline void latency_fsnotify(struct trace_array *tr) { } #endif -- cgit From 0567d6809182df53da03636fad36c507c5cf07a5 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Thu, 14 Nov 2019 14:39:35 -0500 Subject: ftrace: Add modify_ftrace_direct() Add a new function modify_ftrace_direct() that will allow a user to update an existing direct caller to a new trampoline, without missing hits due to unregistering one and then adding another. Link: https://lore.kernel.org/r/20191109022907.6zzo6orhxpt5n2sv@ast-mbp.dhcp.thefacebook.com Suggested-by: Alexei Starovoitov Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/ftrace.c | 78 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 78 insertions(+) (limited to 'kernel') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 82ef8d60a42b..834f3556ea1e 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -5160,6 +5160,84 @@ int unregister_ftrace_direct(unsigned long ip, unsigned long addr) return ret; } EXPORT_SYMBOL_GPL(unregister_ftrace_direct); + +static struct ftrace_ops stub_ops = { + .func = ftrace_stub, +}; + +/** + * modify_ftrace_direct - Modify an existing direct call to call something else + * @ip: The instruction pointer to modify + * @old_addr: The address that the current @ip calls directly + * @new_addr: The address that the @ip should call + * + * This modifies a ftrace direct caller at an instruction pointer without + * having to disable it first. The direct call will switch over to the + * @new_addr without missing anything. + * + * Returns: zero on success. Non zero on error, which includes: + * -ENODEV : the @ip given has no direct caller attached + * -EINVAL : the @old_addr does not match the current direct caller + */ +int modify_ftrace_direct(unsigned long ip, + unsigned long old_addr, unsigned long new_addr) +{ + struct ftrace_func_entry *entry; + struct dyn_ftrace *rec; + int ret = -ENODEV; + + mutex_lock(&direct_mutex); + entry = __ftrace_lookup_ip(direct_functions, ip); + if (!entry) { + /* OK if it is off by a little */ + rec = lookup_rec(ip, ip); + if (!rec || rec->ip == ip) + goto out_unlock; + + entry = __ftrace_lookup_ip(direct_functions, rec->ip); + if (!entry) + goto out_unlock; + + ip = rec->ip; + WARN_ON(!(rec->flags & FTRACE_FL_DIRECT)); + } + + ret = -EINVAL; + if (entry->direct != old_addr) + goto out_unlock; + + /* + * By setting a stub function at the same address, we force + * the code to call the iterator and the direct_ops helper. + * This means that @ip does not call the direct call, and + * we can simply modify it. + */ + ret = ftrace_set_filter_ip(&stub_ops, ip, 0, 0); + if (ret) + goto out_unlock; + + ret = register_ftrace_function(&stub_ops); + if (ret) { + ftrace_set_filter_ip(&stub_ops, ip, 1, 0); + goto out_unlock; + } + + entry->direct = new_addr; + + /* + * By removing the stub, we put back the direct call, calling + * the @new_addr. + */ + unregister_ftrace_function(&stub_ops); + ftrace_set_filter_ip(&stub_ops, ip, 1, 0); + + ret = 0; + + out_unlock: + mutex_unlock(&direct_mutex); + return ret; +} +EXPORT_SYMBOL_GPL(modify_ftrace_direct); #endif /* CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS */ /** -- cgit From 58a74a2925a5b4125dd4f4e728490b9642534c81 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Fri, 15 Nov 2019 11:17:30 +0200 Subject: tracing: Increase SYNTH_FIELDS_MAX for synthetic_events Increase the maximum allowed count of synthetic event fields from 16 to 32 in order to allow for larger-than-usual events. Link: http://lkml.kernel.org/r/20191115091730.9192-1-dedekind1@gmail.com Reviewed-by: Tom Zanussi Signed-off-by: Artem Bityutskiy Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_events_hist.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/trace_events_hist.c b/kernel/trace/trace_events_hist.c index 7482a1466ebf..f49d1a36d3ae 100644 --- a/kernel/trace/trace_events_hist.c +++ b/kernel/trace/trace_events_hist.c @@ -23,7 +23,7 @@ #include "trace_dynevent.h" #define SYNTH_SYSTEM "synthetic" -#define SYNTH_FIELDS_MAX 16 +#define SYNTH_FIELDS_MAX 32 #define STR_VAR_LEN_MAX 32 /* must be multiple of sizeof(u64) */ -- cgit From 1c7f9b673dc0a15753274c4e7f5ebfd4468fc69f Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Fri, 15 Nov 2019 14:13:20 -0500 Subject: ftrace: Fix accounting bug with direct->count in register_ftrace_direct() The direct->count wasn't being updated properly, where it only was updated when the first entry was added, but should be updated every time. Fixes: 013bf0da04748 ("ftrace: Add ftrace_find_direct_func()") Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/ftrace.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 834f3556ea1e..32e4e5ffdd97 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -5093,8 +5093,7 @@ int register_ftrace_direct(unsigned long ip, unsigned long addr) ftrace_direct_func_count--; } } else { - if (!direct->count) - direct->count++; + direct->count++; } out_unlock: mutex_unlock(&direct_mutex); -- cgit From 406acdd32d3e7d5a6dcb7f67798e89068fbe0d77 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Fri, 15 Nov 2019 14:19:04 -0500 Subject: ftrace: Add another check for match in register_ftrace_direct() As an instruction pointer passed into register_ftrace_direct() may just exist on the ftrace call site, but may not be the start of the call site itself, register_ftrace_direct() still needs to update test if a direct call exists on the normalized site, as only one direct call is allowed at any one time. Fixes: 763e34e74bb7d ("ftrace: Add register_ftrace_direct()") Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/ftrace.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 32e4e5ffdd97..9fe33ebaf914 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -5030,7 +5030,12 @@ int register_ftrace_direct(unsigned long ip, unsigned long addr) goto out_unlock; /* Make sure the ip points to the exact record */ - ip = rec->ip; + if (ip != rec->ip) { + ip = rec->ip; + /* Need to check this ip for a direct. */ + if (find_rec_direct(ip)) + goto out_unlock; + } ret = -ENOMEM; if (ftrace_hash_empty(direct_functions) || -- cgit From 128161f47bc3797b0d068da13e311770685d6e4f Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Fri, 15 Nov 2019 14:14:45 -0500 Subject: ftrace: Add helper find_direct_entry() to consolidate code Both unregister_ftrace_direct() and modify_ftrace_direct() needs to normalize the ip passed in to match the rec->ip, as it is acceptable to have the ip on the ftrace call site but not the start. There are also common validity checks with the record found by the ip, these should be done for both unregister_ftrace_direct() and modify_ftrace_direct(). Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/ftrace.c | 59 +++++++++++++++++++++++++-------------------------- 1 file changed, 29 insertions(+), 30 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 9fe33ebaf914..ef79c8393f53 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -5112,30 +5112,40 @@ int register_ftrace_direct(unsigned long ip, unsigned long addr) } EXPORT_SYMBOL_GPL(register_ftrace_direct); -int unregister_ftrace_direct(unsigned long ip, unsigned long addr) +static struct ftrace_func_entry *find_direct_entry(unsigned long *ip) { struct ftrace_func_entry *entry; - struct ftrace_direct_func *direct; struct dyn_ftrace *rec; - int ret = -ENODEV; - mutex_lock(&direct_mutex); + rec = lookup_rec(*ip, *ip); + if (!rec) + return NULL; - entry = __ftrace_lookup_ip(direct_functions, ip); + entry = __ftrace_lookup_ip(direct_functions, rec->ip); if (!entry) { - /* OK if it is off by a little */ - rec = lookup_rec(ip, ip); - if (!rec || rec->ip == ip) - goto out_unlock; + WARN_ON(rec->flags & FTRACE_FL_DIRECT); + return NULL; + } - entry = __ftrace_lookup_ip(direct_functions, rec->ip); - if (!entry) { - WARN_ON(rec->flags & FTRACE_FL_DIRECT); - goto out_unlock; - } + WARN_ON(!(rec->flags & FTRACE_FL_DIRECT)); - WARN_ON(!(rec->flags & FTRACE_FL_DIRECT)); - } + /* Passed in ip just needs to be on the call site */ + *ip = rec->ip; + + return entry; +} + +int unregister_ftrace_direct(unsigned long ip, unsigned long addr) +{ + struct ftrace_direct_func *direct; + struct ftrace_func_entry *entry; + int ret = -ENODEV; + + mutex_lock(&direct_mutex); + + entry = find_direct_entry(&ip); + if (!entry) + goto out_unlock; if (direct_functions->count == 1) unregister_ftrace_function(&direct_ops); @@ -5187,24 +5197,13 @@ int modify_ftrace_direct(unsigned long ip, unsigned long old_addr, unsigned long new_addr) { struct ftrace_func_entry *entry; - struct dyn_ftrace *rec; int ret = -ENODEV; mutex_lock(&direct_mutex); - entry = __ftrace_lookup_ip(direct_functions, ip); - if (!entry) { - /* OK if it is off by a little */ - rec = lookup_rec(ip, ip); - if (!rec || rec->ip == ip) - goto out_unlock; - - entry = __ftrace_lookup_ip(direct_functions, rec->ip); - if (!entry) - goto out_unlock; - ip = rec->ip; - WARN_ON(!(rec->flags & FTRACE_FL_DIRECT)); - } + entry = find_direct_entry(&ip); + if (!entry) + goto out_unlock; ret = -EINVAL; if (entry->direct != old_addr) -- cgit From ea806eb3eab35528b578a061b2c4b28f0f92c465 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Sun, 17 Nov 2019 17:04:15 -0500 Subject: ftrace: Add a helper function to modify_ftrace_direct() to allow arch optimization If a direct ftrace callback is at a location that does not have any other ftrace helpers attached to it, it is possible to simply just change the text to call the new caller (if the architecture supports it). But this requires special architecture code. Currently, modify_ftrace_direct() uses a trick to add a stub ftrace callback to the location forcing it to call the ftrace iterator. Then it can change the direct helper to call the new function in C, and then remove the stub. Removing the stub will have the location now call the new location that the direct helper is using. The new helper function does the registering the stub trick, but is a weak function, allowing an architecture to override it to do something a bit more direct. Link: https://lore.kernel.org/r/20191115215125.mbqv7taqnx376yed@ast-mbp.dhcp.thefacebook.com Suggested-by: Alexei Starovoitov Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/ftrace.c | 120 ++++++++++++++++++++++++++++++++++++-------------- 1 file changed, 88 insertions(+), 32 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index ef79c8393f53..caae523f4ef3 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1020,12 +1020,6 @@ static bool update_all_ops; # error Dynamic ftrace depends on MCOUNT_RECORD #endif -struct ftrace_func_entry { - struct hlist_node hlist; - unsigned long ip; - unsigned long direct; /* for direct lookup only */ -}; - struct ftrace_func_probe { struct ftrace_probe_ops *probe_ops; struct ftrace_ops ops; @@ -5112,7 +5106,8 @@ int register_ftrace_direct(unsigned long ip, unsigned long addr) } EXPORT_SYMBOL_GPL(register_ftrace_direct); -static struct ftrace_func_entry *find_direct_entry(unsigned long *ip) +static struct ftrace_func_entry *find_direct_entry(unsigned long *ip, + struct dyn_ftrace **recp) { struct ftrace_func_entry *entry; struct dyn_ftrace *rec; @@ -5132,6 +5127,9 @@ static struct ftrace_func_entry *find_direct_entry(unsigned long *ip) /* Passed in ip just needs to be on the call site */ *ip = rec->ip; + if (recp) + *recp = rec; + return entry; } @@ -5143,7 +5141,7 @@ int unregister_ftrace_direct(unsigned long ip, unsigned long addr) mutex_lock(&direct_mutex); - entry = find_direct_entry(&ip); + entry = find_direct_entry(&ip, NULL); if (!entry) goto out_unlock; @@ -5179,6 +5177,75 @@ static struct ftrace_ops stub_ops = { .func = ftrace_stub, }; +/** + * ftrace_modify_direct_caller - modify ftrace nop directly + * @entry: The ftrace hash entry of the direct helper for @rec + * @rec: The record representing the function site to patch + * @old_addr: The location that the site at @rec->ip currently calls + * @new_addr: The location that the site at @rec->ip should call + * + * An architecture may overwrite this function to optimize the + * changing of the direct callback on an ftrace nop location. + * This is called with the ftrace_lock mutex held, and no other + * ftrace callbacks are on the associated record (@rec). Thus, + * it is safe to modify the ftrace record, where it should be + * currently calling @old_addr directly, to call @new_addr. + * + * Safety checks should be made to make sure that the code at + * @rec->ip is currently calling @old_addr. And this must + * also update entry->direct to @new_addr. + */ +int __weak ftrace_modify_direct_caller(struct ftrace_func_entry *entry, + struct dyn_ftrace *rec, + unsigned long old_addr, + unsigned long new_addr) +{ + unsigned long ip = rec->ip; + int ret; + + /* + * The ftrace_lock was used to determine if the record + * had more than one registered user to it. If it did, + * we needed to prevent that from changing to do the quick + * switch. But if it did not (only a direct caller was attached) + * then this function is called. But this function can deal + * with attached callers to the rec that we care about, and + * since this function uses standard ftrace calls that take + * the ftrace_lock mutex, we need to release it. + */ + mutex_unlock(&ftrace_lock); + + /* + * By setting a stub function at the same address, we force + * the code to call the iterator and the direct_ops helper. + * This means that @ip does not call the direct call, and + * we can simply modify it. + */ + ret = ftrace_set_filter_ip(&stub_ops, ip, 0, 0); + if (ret) + goto out_lock; + + ret = register_ftrace_function(&stub_ops); + if (ret) { + ftrace_set_filter_ip(&stub_ops, ip, 1, 0); + goto out_lock; + } + + entry->direct = new_addr; + + /* + * By removing the stub, we put back the direct call, calling + * the @new_addr. + */ + unregister_ftrace_function(&stub_ops); + ftrace_set_filter_ip(&stub_ops, ip, 1, 0); + + out_lock: + mutex_lock(&ftrace_lock); + + return ret; +} + /** * modify_ftrace_direct - Modify an existing direct call to call something else * @ip: The instruction pointer to modify @@ -5197,11 +5264,13 @@ int modify_ftrace_direct(unsigned long ip, unsigned long old_addr, unsigned long new_addr) { struct ftrace_func_entry *entry; + struct dyn_ftrace *rec; int ret = -ENODEV; mutex_lock(&direct_mutex); - entry = find_direct_entry(&ip); + mutex_lock(&ftrace_lock); + entry = find_direct_entry(&ip, &rec); if (!entry) goto out_unlock; @@ -5210,33 +5279,20 @@ int modify_ftrace_direct(unsigned long ip, goto out_unlock; /* - * By setting a stub function at the same address, we force - * the code to call the iterator and the direct_ops helper. - * This means that @ip does not call the direct call, and - * we can simply modify it. + * If there's no other ftrace callback on the rec->ip location, + * then it can be changed directly by the architecture. + * If there is another caller, then we just need to change the + * direct caller helper to point to @new_addr. */ - ret = ftrace_set_filter_ip(&stub_ops, ip, 0, 0); - if (ret) - goto out_unlock; - - ret = register_ftrace_function(&stub_ops); - if (ret) { - ftrace_set_filter_ip(&stub_ops, ip, 1, 0); - goto out_unlock; + if (ftrace_rec_count(rec) == 1) { + ret = ftrace_modify_direct_caller(entry, rec, old_addr, new_addr); + } else { + entry->direct = new_addr; + ret = 0; } - entry->direct = new_addr; - - /* - * By removing the stub, we put back the direct call, calling - * the @new_addr. - */ - unregister_ftrace_function(&stub_ops); - ftrace_set_filter_ip(&stub_ops, ip, 1, 0); - - ret = 0; - out_unlock: + mutex_unlock(&ftrace_lock); mutex_unlock(&direct_mutex); return ret; } -- cgit From 46f9469247c6f4697cbbf37e4b3961120bf07f29 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Mon, 18 Nov 2019 10:41:29 -0500 Subject: ftrace: Rename ftrace_graph_stub to ftrace_stub_graph The ftrace_graph_stub was created and points to ftrace_stub as a way to assign the functon graph tracer function pointer to a stub function with a different prototype than what ftrace_stub has and not trigger the C verifier. The ftrace_graph_stub was created via the linker script vmlinux.lds.h. Unfortunately, powerpc already uses the name ftrace_graph_stub for its internal implementation of the function graph tracer, and even though powerpc would still build, the change via the linker script broke function tracer on powerpc from working. By using the name ftrace_stub_graph, which does not exist anywhere else in the kernel, this should not be a problem. Link: https://lore.kernel.org/r/1573849732.5937.136.camel@lca.pw Fixes: b83b43ffc6e4 ("fgraph: Fix function type mismatches of ftrace_graph_return using ftrace_stub") Reorted-by: Qian Cai Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/fgraph.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/fgraph.c b/kernel/trace/fgraph.c index fa3ce10d0405..67e0c462b059 100644 --- a/kernel/trace/fgraph.c +++ b/kernel/trace/fgraph.c @@ -336,10 +336,10 @@ int ftrace_graph_entry_stub(struct ftrace_graph_ent *trace) * Simply points to ftrace_stub, but with the proper protocol. * Defined by the linker script in linux/vmlinux.lds.h */ -extern void ftrace_graph_stub(struct ftrace_graph_ret *); +extern void ftrace_stub_graph(struct ftrace_graph_ret *); /* The callbacks that hook a function */ -trace_func_graph_ret_t ftrace_graph_return = ftrace_graph_stub; +trace_func_graph_ret_t ftrace_graph_return = ftrace_stub_graph; trace_func_graph_ent_t ftrace_graph_entry = ftrace_graph_entry_stub; static trace_func_graph_ent_t __ftrace_graph_entry = ftrace_graph_entry_stub; @@ -619,7 +619,7 @@ void unregister_ftrace_graph(struct fgraph_ops *gops) goto out; ftrace_graph_active--; - ftrace_graph_return = ftrace_graph_stub; + ftrace_graph_return = ftrace_stub_graph; ftrace_graph_entry = ftrace_graph_entry_stub; __ftrace_graph_entry = ftrace_graph_entry_stub; ftrace_shutdown(&graph_ops, FTRACE_STOP_FUNC_RET); -- cgit From a82a4804b4ee3636c8988fea14d44f70f4de45f1 Mon Sep 17 00:00:00 2001 From: Xianting Tian Date: Sat, 16 Nov 2019 10:05:55 -0500 Subject: ring-buffer: Fix typos in function ring_buffer_producer Fix spelling and other typos Link: http://lkml.kernel.org/r/1573916755-32478-1-git-send-email-xianting_tian@126.com Signed-off-by: Xianting Tian Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/ring_buffer_benchmark.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/ring_buffer_benchmark.c b/kernel/trace/ring_buffer_benchmark.c index 09b0b49f346e..32149e46551c 100644 --- a/kernel/trace/ring_buffer_benchmark.c +++ b/kernel/trace/ring_buffer_benchmark.c @@ -269,10 +269,10 @@ static void ring_buffer_producer(void) #ifndef CONFIG_PREEMPTION /* - * If we are a non preempt kernel, the 10 second run will + * If we are a non preempt kernel, the 10 seconds run will * stop everything while it runs. Instead, we will call * cond_resched and also add any time that was lost by a - * rescedule. + * reschedule. * * Do a cond resched at the same frequency we would wake up * the reader. -- cgit From fc809bc5ceaa665497384064ab2d76713c774bad Mon Sep 17 00:00:00 2001 From: Krzysztof Kozlowski Date: Wed, 20 Nov 2019 21:38:07 +0800 Subject: tracing: Fix Kconfig indentation Adjust indentation from spaces to tab (+optional two spaces) as in coding style with command like: $ sed -e 's/^ /\t/' -i */Kconfig Link: http://lkml.kernel.org/r/20191120133807.12741-1-krzk@kernel.org Signed-off-by: Krzysztof Kozlowski Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/Kconfig | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index b872716bb2a0..f67620499faa 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -79,7 +79,7 @@ config FTRACE_NMI_ENTER config EVENT_TRACING select CONTEXT_SWITCH_TRACER - select GLOB + select GLOB bool config CONTEXT_SWITCH_TRACER @@ -311,7 +311,7 @@ config TRACER_SNAPSHOT cat snapshot config TRACER_SNAPSHOT_PER_CPU_SWAP - bool "Allow snapshot to swap per CPU" + bool "Allow snapshot to swap per CPU" depends on TRACER_SNAPSHOT select RING_BUFFER_ALLOW_SWAP help @@ -683,7 +683,7 @@ config MMIOTRACE_TEST Say N, unless you absolutely know what you are doing. config TRACEPOINT_BENCHMARK - bool "Add tracepoint that benchmarks tracepoints" + bool "Add tracepoint that benchmarks tracepoints" help This option creates the tracepoint "benchmark:benchmark_event". When the tracepoint is enabled, it kicks off a kernel thread that @@ -732,7 +732,7 @@ config RING_BUFFER_STARTUP_TEST bool "Ring buffer startup self test" depends on RING_BUFFER help - Run a simple self test on the ring buffer on boot up. Late in the + Run a simple self test on the ring buffer on boot up. Late in the kernel boot sequence, the test will start that kicks off a thread per cpu. Each thread will write various size events into the ring buffer. Another thread is created to send IPIs -- cgit From 28879787147358e8ffcae397f11748de3dd26577 Mon Sep 17 00:00:00 2001 From: Divya Indi Date: Wed, 20 Nov 2019 11:08:38 -0800 Subject: tracing: Adding new functions for kernel access to Ftrace instances Adding 2 new functions - 1) struct trace_array *trace_array_get_by_name(const char *name); Return pointer to a trace array with given name. If it does not exist, create and return pointer to the new trace array. 2) int trace_array_set_clr_event(struct trace_array *tr, const char *system ,const char *event, bool enable); Enable/Disable events to this trace array. Additionally, - To handle reference counters, export trace_array_put() - Due to introduction of the above 2 new functions, we no longer need to export - ftrace_set_clr_event & trace_array_create APIs. Link: http://lkml.kernel.org/r/1574276919-11119-2-git-send-email-divya.indi@oracle.com Signed-off-by: Divya Indi Reviewed-by: Aruna Ramakrishna Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace.c | 96 +++++++++++++++++++++++++++++++++++---------- kernel/trace/trace.h | 1 - kernel/trace/trace_events.c | 27 ++++++++++++- 3 files changed, 102 insertions(+), 22 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 42659ce6ac0c..02a23a6e5e00 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -301,12 +301,24 @@ static void __trace_array_put(struct trace_array *this_tr) this_tr->ref--; } +/** + * trace_array_put - Decrement the reference counter for this trace array. + * + * NOTE: Use this when we no longer need the trace array returned by + * trace_array_get_by_name(). This ensures the trace array can be later + * destroyed. + * + */ void trace_array_put(struct trace_array *this_tr) { + if (!this_tr) + return; + mutex_lock(&trace_types_lock); __trace_array_put(this_tr); mutex_unlock(&trace_types_lock); } +EXPORT_SYMBOL_GPL(trace_array_put); int tracing_check_open_get_tr(struct trace_array *tr) { @@ -8437,24 +8449,15 @@ static void update_tracer_options(struct trace_array *tr) mutex_unlock(&trace_types_lock); } -struct trace_array *trace_array_create(const char *name) +static struct trace_array *trace_array_create(const char *name) { struct trace_array *tr; int ret; - mutex_lock(&event_mutex); - mutex_lock(&trace_types_lock); - - ret = -EEXIST; - list_for_each_entry(tr, &ftrace_trace_arrays, list) { - if (tr->name && strcmp(tr->name, name) == 0) - goto out_unlock; - } - ret = -ENOMEM; tr = kzalloc(sizeof(*tr), GFP_KERNEL); if (!tr) - goto out_unlock; + return ERR_PTR(ret); tr->name = kstrdup(name, GFP_KERNEL); if (!tr->name) @@ -8499,8 +8502,8 @@ struct trace_array *trace_array_create(const char *name) list_add(&tr->list, &ftrace_trace_arrays); - mutex_unlock(&trace_types_lock); - mutex_unlock(&event_mutex); + tr->ref++; + return tr; @@ -8510,24 +8513,77 @@ struct trace_array *trace_array_create(const char *name) kfree(tr->name); kfree(tr); - out_unlock: - mutex_unlock(&trace_types_lock); - mutex_unlock(&event_mutex); - return ERR_PTR(ret); } -EXPORT_SYMBOL_GPL(trace_array_create); static int instance_mkdir(const char *name) { - return PTR_ERR_OR_ZERO(trace_array_create(name)); + struct trace_array *tr; + int ret; + + mutex_lock(&event_mutex); + mutex_lock(&trace_types_lock); + + ret = -EEXIST; + list_for_each_entry(tr, &ftrace_trace_arrays, list) { + if (tr->name && strcmp(tr->name, name) == 0) + goto out_unlock; + } + + tr = trace_array_create(name); + + ret = PTR_ERR_OR_ZERO(tr); + +out_unlock: + mutex_unlock(&trace_types_lock); + mutex_unlock(&event_mutex); + return ret; +} + +/** + * trace_array_get_by_name - Create/Lookup a trace array, given its name. + * @name: The name of the trace array to be looked up/created. + * + * Returns pointer to trace array with given name. + * NULL, if it cannot be created. + * + * NOTE: This function increments the reference counter associated with the + * trace array returned. This makes sure it cannot be freed while in use. + * Use trace_array_put() once the trace array is no longer needed. + * + */ +struct trace_array *trace_array_get_by_name(const char *name) +{ + struct trace_array *tr; + + mutex_lock(&event_mutex); + mutex_lock(&trace_types_lock); + + list_for_each_entry(tr, &ftrace_trace_arrays, list) { + if (tr->name && strcmp(tr->name, name) == 0) + goto out_unlock; + } + + tr = trace_array_create(name); + + if (IS_ERR(tr)) + tr = NULL; +out_unlock: + if (tr) + tr->ref++; + + mutex_unlock(&trace_types_lock); + mutex_unlock(&event_mutex); + return tr; } +EXPORT_SYMBOL_GPL(trace_array_get_by_name); static int __remove_instance(struct trace_array *tr) { int i; - if (tr->ref || (tr->current_trace && tr->current_trace->ref)) + /* Reference counter for a newly created trace array = 1. */ + if (tr->ref > 1 || (tr->current_trace && tr->current_trace->ref)) return -EBUSY; list_del(&tr->list); diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 2df8aed6a8f0..ca7fccafbcbb 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -345,7 +345,6 @@ extern struct list_head ftrace_trace_arrays; extern struct mutex trace_types_lock; extern int trace_array_get(struct trace_array *tr); -extern void trace_array_put(struct trace_array *tr); extern int tracing_check_open_get_tr(struct trace_array *tr); extern int tracing_set_time_stamp_abs(struct trace_array *tr, bool abs); diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 2a3ac2365445..6b3a69e9aa6a 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -827,7 +827,6 @@ int ftrace_set_clr_event(struct trace_array *tr, char *buf, int set) return ret; } -EXPORT_SYMBOL_GPL(ftrace_set_clr_event); /** * trace_set_clr_event - enable or disable an event @@ -852,6 +851,32 @@ int trace_set_clr_event(const char *system, const char *event, int set) } EXPORT_SYMBOL_GPL(trace_set_clr_event); +/** + * trace_array_set_clr_event - enable or disable an event for a trace array. + * @tr: concerned trace array. + * @system: system name to match (NULL for any system) + * @event: event name to match (NULL for all events, within system) + * @enable: true to enable, false to disable + * + * This is a way for other parts of the kernel to enable or disable + * event recording. + * + * Returns 0 on success, -EINVAL if the parameters do not match any + * registered events. + */ +int trace_array_set_clr_event(struct trace_array *tr, const char *system, + const char *event, bool enable) +{ + int set; + + if (!tr) + return -ENOENT; + + set = (enable == true) ? 1 : 0; + return __ftrace_set_clr_event(tr, NULL, system, event, set); +} +EXPORT_SYMBOL_GPL(trace_array_set_clr_event); + /* 128 should be much more than enough */ #define EVENT_BUF_SIZE 127 -- cgit From 0e24220821b0e0e330a18bfef29ac6396545d62e Mon Sep 17 00:00:00 2001 From: Hassan Naveed Date: Fri, 15 Nov 2019 23:44:42 +0000 Subject: tracing: Use xarray for syscall trace events Currently, a lot of memory is wasted for architectures like MIPS when init_ftrace_syscalls() allocates the array for syscalls using kcalloc. This is because syscalls numbers start from 4000, 5000 or 6000 and array elements up to that point are unused. Fix this by using a data structure more suited to storing sparsely populated arrays. The XARRAY data structure, implemented using radix trees, is much more memory efficient for storing the syscalls in question. Link: http://lkml.kernel.org/r/20191115234314.21599-1-hnaveed@wavecomp.com Signed-off-by: Hassan Naveed Reviewed-by: Paul Burton Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_syscalls.c | 32 +++++++++++++++++++++++++------- 1 file changed, 25 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index fa8fbff736d6..16fa218556fa 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -7,6 +7,7 @@ #include /* for MODULE_NAME_LEN via KSYM_SYMBOL_LEN */ #include #include +#include #include #include "trace_output.h" @@ -30,6 +31,7 @@ syscall_get_enter_fields(struct trace_event_call *call) extern struct syscall_metadata *__start_syscalls_metadata[]; extern struct syscall_metadata *__stop_syscalls_metadata[]; +static DEFINE_XARRAY(syscalls_metadata_sparse); static struct syscall_metadata **syscalls_metadata; #ifndef ARCH_HAS_SYSCALL_MATCH_SYM_NAME @@ -101,6 +103,9 @@ find_syscall_meta(unsigned long syscall) static struct syscall_metadata *syscall_nr_to_meta(int nr) { + if (IS_ENABLED(CONFIG_HAVE_SPARSE_SYSCALL_NR)) + return xa_load(&syscalls_metadata_sparse, (unsigned long)nr); + if (!syscalls_metadata || nr >= NR_syscalls || nr < 0) return NULL; @@ -536,12 +541,16 @@ void __init init_ftrace_syscalls(void) struct syscall_metadata *meta; unsigned long addr; int i; - - syscalls_metadata = kcalloc(NR_syscalls, sizeof(*syscalls_metadata), - GFP_KERNEL); - if (!syscalls_metadata) { - WARN_ON(1); - return; + void *ret; + + if (!IS_ENABLED(CONFIG_HAVE_SPARSE_SYSCALL_NR)) { + syscalls_metadata = kcalloc(NR_syscalls, + sizeof(*syscalls_metadata), + GFP_KERNEL); + if (!syscalls_metadata) { + WARN_ON(1); + return; + } } for (i = 0; i < NR_syscalls; i++) { @@ -551,7 +560,16 @@ void __init init_ftrace_syscalls(void) continue; meta->syscall_nr = i; - syscalls_metadata[i] = meta; + + if (!IS_ENABLED(CONFIG_HAVE_SPARSE_SYSCALL_NR)) { + syscalls_metadata[i] = meta; + } else { + ret = xa_store(&syscalls_metadata_sparse, i, meta, + GFP_KERNEL); + WARN(xa_is_err(ret), + "Syscall memory allocation failed\n"); + } + } } -- cgit