AuxXxilium/linux_dsm_epyc7002
1
0
Fork 0
mirror of https://github.com/AuxXxilium/linux_dsm_epyc7002.git synced 2024-11-24 03:40:52 +07:00
Code Issues Actions Packages Projects Releases Wiki Activity

kcov: collect coverage from interrupts

Browse Source 
This change extends kcov remote coverage support to allow collecting
coverage from soft interrupts in addition to kernel background threads.

To collect coverage from code that is executed in softirq context, a part
of that code has to be annotated with kcov_remote_start/stop() in a
similar way as how it is done for global kernel background threads.  Then
the handle used for the annotations has to be passed to the
KCOV_REMOTE_ENABLE ioctl.

Internally this patch adjusts the __sanitizer_cov_trace_pc() compiler
inserted callback to not bail out when called from softirq context.
kcov_remote_start/stop() are updated to save/restore the current per task
kcov state in a per-cpu area (in case the softirq came when the kernel was
already collecting coverage in task context).  Coverage from softirqs is
collected into pre-allocated per-cpu areas, whose size is controlled by
the new CONFIG_KCOV_IRQ_AREA_SIZE.

[andreyknvl@google.com: turn current->kcov_softirq into unsigned int to fix objtool warning]
  Link: http://lkml.kernel.org/r/841c778aa3849c5cb8c3761f56b87ce653a88671.1585233617.git.andreyknvl@google.com
Signed-off-by: Andrey Konovalov <andreyknvl@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Dmitry Vyukov <dvyukov@google.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Alexander Potapenko <glider@google.com>
Cc: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Marco Elver <elver@google.com>
Link: http://lkml.kernel.org/r/469bd385c431d050bc38a593296eff4baae50666.1584655448.git.andreyknvl@google.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This commit is contained in:
Andrey Konovalov 2020-06-04 16:46:04 -07:00 committed by Linus Torvalds
parent 5fe7042dc0
commit 5ff3b30ab5
4 changed files with 176 additions and 47 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
Documentation/dev-tools/kcov.rst
View File

@ -217,14 +217,15 @@ This allows to collect coverage from two types of kernel background
threads: the global ones, that are spawned during kernel boot in a limited threads: the global ones, that are spawned during kernel boot in a limited
number of instances (e.g. one USB hub_event() worker thread is spawned per number of instances (e.g. one USB hub_event() worker thread is spawned per
USB HCD); and the local ones, that are spawned when a user interacts with USB HCD); and the local ones, that are spawned when a user interacts with
some kernel interface (e.g. vhost workers). some kernel interface (e.g. vhost workers); as well as from soft
interrupts.
To enable collecting coverage from a global background thread, a unique To enable collecting coverage from a global background thread or from a
global handle must be assigned and passed to the corresponding softirq, a unique global handle must be assigned and passed to the
kcov_remote_start() call. Then a userspace process can pass a list of such corresponding kcov_remote_start() call. Then a userspace process can pass
handles to the KCOV_REMOTE_ENABLE ioctl in the handles array field of the a list of such handles to the KCOV_REMOTE_ENABLE ioctl in the handles
kcov_remote_arg struct. This will attach the used kcov device to the code array field of the kcov_remote_arg struct. This will attach the used kcov
sections, that are referenced by those handles. device to the code sections, that are referenced by those handles.
Since there might be many local background threads spawned from different Since there might be many local background threads spawned from different
userspace processes, we can't use a single global handle per annotation. userspace processes, we can't use a single global handle per annotation.
@ -242,7 +243,7 @@ handles as they don't belong to a particular subsystem. The bytes 4-7 are
currently reserved and must be zero. In the future the number of bytes currently reserved and must be zero. In the future the number of bytes
used for the subsystem or handle ids might be increased. used for the subsystem or handle ids might be increased.
When a particular userspace proccess collects coverage by via a common When a particular userspace proccess collects coverage via a common
handle, kcov will collect coverage for each code section that is annotated handle, kcov will collect coverage for each code section that is annotated
to use the common handle obtained as kcov_handle from the current to use the common handle obtained as kcov_handle from the current
task_struct. However non common handles allow to collect coverage task_struct. However non common handles allow to collect coverage

3
include/linux/sched.h
View File

@ -1247,6 +1247,9 @@ struct task_struct {
/* KCOV sequence number: */ /* KCOV sequence number: */
int kcov_sequence; int kcov_sequence;
/* Collect coverage from softirq context: */
unsigned int kcov_softirq;
#endif #endif
#ifdef CONFIG_MEMCG #ifdef CONFIG_MEMCG

194
kernel/kcov.c
View File

@ -86,6 +86,18 @@ static DEFINE_SPINLOCK(kcov_remote_lock);
static DEFINE_HASHTABLE(kcov_remote_map, 4); static DEFINE_HASHTABLE(kcov_remote_map, 4);
static struct list_head kcov_remote_areas = LIST_HEAD_INIT(kcov_remote_areas); static struct list_head kcov_remote_areas = LIST_HEAD_INIT(kcov_remote_areas);
struct kcov_percpu_data {
void *irq_area;
unsigned int saved_mode;
unsigned int saved_size;
void *saved_area;
struct kcov *saved_kcov;
int saved_sequence;
};
DEFINE_PER_CPU(struct kcov_percpu_data, kcov_percpu_data);
/* Must be called with kcov_remote_lock locked. */ /* Must be called with kcov_remote_lock locked. */
static struct kcov_remote *kcov_remote_find(u64 handle) static struct kcov_remote *kcov_remote_find(u64 handle)
{ {
@ -145,9 +157,10 @@ static notrace bool check_kcov_mode(enum kcov_mode needed_mode, struct task_stru
/* /*
* We are interested in code coverage as a function of a syscall inputs, * We are interested in code coverage as a function of a syscall inputs,
* so we ignore code executed in interrupts. * so we ignore code executed in interrupts, unless we are in a remote
* coverage collection section in a softirq.
*/ */
if (!in_task()) if (!in_task() && !(in_serving_softirq() && t->kcov_softirq))
return false; return false;
mode = READ_ONCE(t->kcov_mode); mode = READ_ONCE(t->kcov_mode);
/* /*
@ -360,8 +373,9 @@ static void kcov_remote_reset(struct kcov *kcov)
int bkt; int bkt;
struct kcov_remote *remote; struct kcov_remote *remote;
struct hlist_node *tmp; struct hlist_node *tmp;
unsigned long flags;
spin_lock(&kcov_remote_lock); spin_lock_irqsave(&kcov_remote_lock, flags);
hash_for_each_safe(kcov_remote_map, bkt, tmp, remote, hnode) { hash_for_each_safe(kcov_remote_map, bkt, tmp, remote, hnode) {
if (remote->kcov != kcov) if (remote->kcov != kcov)
continue; continue;
@ -370,7 +384,7 @@ static void kcov_remote_reset(struct kcov *kcov)
} }
/* Do reset before unlock to prevent races with kcov_remote_start(). */ /* Do reset before unlock to prevent races with kcov_remote_start(). */
kcov_reset(kcov); kcov_reset(kcov);
spin_unlock(&kcov_remote_lock); spin_unlock_irqrestore(&kcov_remote_lock, flags);
} }
static void kcov_disable(struct task_struct *t, struct kcov *kcov) static void kcov_disable(struct task_struct *t, struct kcov *kcov)
@ -399,12 +413,13 @@ static void kcov_put(struct kcov *kcov)
void kcov_task_exit(struct task_struct *t) void kcov_task_exit(struct task_struct *t)
{ {
struct kcov *kcov; struct kcov *kcov;
unsigned long flags;
kcov = t->kcov; kcov = t->kcov;
if (kcov == NULL) if (kcov == NULL)
return; return;
spin_lock(&kcov->lock); spin_lock_irqsave(&kcov->lock, flags);
kcov_debug("t = %px, kcov->t = %px\n", t, kcov->t); kcov_debug("t = %px, kcov->t = %px\n", t, kcov->t);
/* /*
* For KCOV_ENABLE devices we want to make sure that t->kcov->t == t, * For KCOV_ENABLE devices we want to make sure that t->kcov->t == t,
@ -428,12 +443,12 @@ void kcov_task_exit(struct task_struct *t)
* By combining all three checks into one we get: * By combining all three checks into one we get:
*/ */
if (WARN_ON(kcov->t != t)) { if (WARN_ON(kcov->t != t)) {
spin_unlock(&kcov->lock); spin_unlock_irqrestore(&kcov->lock, flags);
return; return;
} }
/* Just to not leave dangling references behind. */ /* Just to not leave dangling references behind. */
kcov_disable(t, kcov); kcov_disable(t, kcov);
spin_unlock(&kcov->lock); spin_unlock_irqrestore(&kcov->lock, flags);
kcov_put(kcov); kcov_put(kcov);
} }
@ -444,12 +459,13 @@ static int kcov_mmap(struct file *filep, struct vm_area_struct *vma)
struct kcov *kcov = vma->vm_file->private_data; struct kcov *kcov = vma->vm_file->private_data;
unsigned long size, off; unsigned long size, off;
struct page *page; struct page *page;
unsigned long flags;
area = vmalloc_user(vma->vm_end - vma->vm_start); area = vmalloc_user(vma->vm_end - vma->vm_start);
if (!area) if (!area)
return -ENOMEM; return -ENOMEM;
spin_lock(&kcov->lock); spin_lock_irqsave(&kcov->lock, flags);
size = kcov->size * sizeof(unsigned long); size = kcov->size * sizeof(unsigned long);
if (kcov->mode != KCOV_MODE_INIT || vma->vm_pgoff != 0 || if (kcov->mode != KCOV_MODE_INIT || vma->vm_pgoff != 0 ||
vma->vm_end - vma->vm_start != size) { vma->vm_end - vma->vm_start != size) {
@ -459,7 +475,7 @@ static int kcov_mmap(struct file *filep, struct vm_area_struct *vma)
if (!kcov->area) { if (!kcov->area) {
kcov->area = area; kcov->area = area;
vma->vm_flags |= VM_DONTEXPAND; vma->vm_flags |= VM_DONTEXPAND;
spin_unlock(&kcov->lock); spin_unlock_irqrestore(&kcov->lock, flags);
for (off = 0; off < size; off += PAGE_SIZE) { for (off = 0; off < size; off += PAGE_SIZE) {
page = vmalloc_to_page(kcov->area + off); page = vmalloc_to_page(kcov->area + off);
if (vm_insert_page(vma, vma->vm_start + off, page)) if (vm_insert_page(vma, vma->vm_start + off, page))
@ -468,7 +484,7 @@ static int kcov_mmap(struct file *filep, struct vm_area_struct *vma)
return 0; return 0;
} }
exit: exit:
spin_unlock(&kcov->lock); spin_unlock_irqrestore(&kcov->lock, flags);
vfree(area); vfree(area);
return res; return res;
} }
@ -548,6 +564,7 @@ static int kcov_ioctl_locked(struct kcov *kcov, unsigned int cmd,
int mode, i; int mode, i;
struct kcov_remote_arg *remote_arg; struct kcov_remote_arg *remote_arg;
struct kcov_remote *remote; struct kcov_remote *remote;
unsigned long flags;
switch (cmd) { switch (cmd) {
case KCOV_INIT_TRACE: case KCOV_INIT_TRACE:
@ -620,17 +637,19 @@ static int kcov_ioctl_locked(struct kcov *kcov, unsigned int cmd,
kcov->t = t; kcov->t = t;
kcov->remote = true; kcov->remote = true;
kcov->remote_size = remote_arg->area_size; kcov->remote_size = remote_arg->area_size;
spin_lock(&kcov_remote_lock); spin_lock_irqsave(&kcov_remote_lock, flags);
for (i = 0; i < remote_arg->num_handles; i++) { for (i = 0; i < remote_arg->num_handles; i++) {
if (!kcov_check_handle(remote_arg->handles[i], if (!kcov_check_handle(remote_arg->handles[i],
false, true, false)) { false, true, false)) {
spin_unlock(&kcov_remote_lock); spin_unlock_irqrestore(&kcov_remote_lock,
flags);
kcov_disable(t, kcov); kcov_disable(t, kcov);
return -EINVAL; return -EINVAL;
} }
remote = kcov_remote_add(kcov, remote_arg->handles[i]); remote = kcov_remote_add(kcov, remote_arg->handles[i]);
if (IS_ERR(remote)) { if (IS_ERR(remote)) {
spin_unlock(&kcov_remote_lock); spin_unlock_irqrestore(&kcov_remote_lock,
flags);
kcov_disable(t, kcov); kcov_disable(t, kcov);
return PTR_ERR(remote); return PTR_ERR(remote);
} }
@ -638,20 +657,22 @@ static int kcov_ioctl_locked(struct kcov *kcov, unsigned int cmd,
if (remote_arg->common_handle) { if (remote_arg->common_handle) {
if (!kcov_check_handle(remote_arg->common_handle, if (!kcov_check_handle(remote_arg->common_handle,
true, false, false)) { true, false, false)) {
spin_unlock(&kcov_remote_lock); spin_unlock_irqrestore(&kcov_remote_lock,
flags);
kcov_disable(t, kcov); kcov_disable(t, kcov);
return -EINVAL; return -EINVAL;
} }
remote = kcov_remote_add(kcov, remote = kcov_remote_add(kcov,
remote_arg->common_handle); remote_arg->common_handle);
if (IS_ERR(remote)) { if (IS_ERR(remote)) {
spin_unlock(&kcov_remote_lock); spin_unlock_irqrestore(&kcov_remote_lock,
flags);
kcov_disable(t, kcov); kcov_disable(t, kcov);
return PTR_ERR(remote); return PTR_ERR(remote);
} }
t->kcov_handle = remote_arg->common_handle; t->kcov_handle = remote_arg->common_handle;
} }
spin_unlock(&kcov_remote_lock); spin_unlock_irqrestore(&kcov_remote_lock, flags);
/* Put either in kcov_task_exit() or in KCOV_DISABLE. */ /* Put either in kcov_task_exit() or in KCOV_DISABLE. */
kcov_get(kcov); kcov_get(kcov);
return 0; return 0;
@ -667,6 +688,7 @@ static long kcov_ioctl(struct file *filep, unsigned int cmd, unsigned long arg)
struct kcov_remote_arg *remote_arg = NULL; struct kcov_remote_arg *remote_arg = NULL;
unsigned int remote_num_handles; unsigned int remote_num_handles;
unsigned long remote_arg_size; unsigned long remote_arg_size;
unsigned long flags;
if (cmd == KCOV_REMOTE_ENABLE) { if (cmd == KCOV_REMOTE_ENABLE) {
if (get_user(remote_num_handles, (unsigned __user *)(arg + if (get_user(remote_num_handles, (unsigned __user *)(arg +
@ -687,9 +709,9 @@ static long kcov_ioctl(struct file *filep, unsigned int cmd, unsigned long arg)
} }
kcov = filep->private_data; kcov = filep->private_data;
spin_lock(&kcov->lock); spin_lock_irqsave(&kcov->lock, flags);
res = kcov_ioctl_locked(kcov, cmd, arg); res = kcov_ioctl_locked(kcov, cmd, arg);
spin_unlock(&kcov->lock); spin_unlock_irqrestore(&kcov->lock, flags);
kfree(remote_arg); kfree(remote_arg);
@ -706,8 +728,8 @@ static const struct file_operations kcov_fops = {
/* /*
* kcov_remote_start() and kcov_remote_stop() can be used to annotate a section * kcov_remote_start() and kcov_remote_stop() can be used to annotate a section
* of code in a kernel background thread to allow kcov to be used to collect * of code in a kernel background thread or in a softirq to allow kcov to be
* coverage from that part of code. * used to collect coverage from that part of code.
* *
* The handle argument of kcov_remote_start() identifies a code section that is * The handle argument of kcov_remote_start() identifies a code section that is
* used for coverage collection. A userspace process passes this handle to * used for coverage collection. A userspace process passes this handle to
@ -718,9 +740,9 @@ static const struct file_operations kcov_fops = {
* the type of the kernel thread whose code is being annotated. * the type of the kernel thread whose code is being annotated.
* *
* For global kernel threads that are spawned in a limited number of instances * For global kernel threads that are spawned in a limited number of instances
* (e.g. one USB hub_event() worker thread is spawned per USB HCD), each * (e.g. one USB hub_event() worker thread is spawned per USB HCD) and for
* instance must be assigned a unique 4-byte instance id. The instance id is * softirqs, each instance must be assigned a unique 4-byte instance id. The
* then combined with a 1-byte subsystem id to get a handle via * instance id is then combined with a 1-byte subsystem id to get a handle via
* kcov_remote_handle(subsystem_id, instance_id). * kcov_remote_handle(subsystem_id, instance_id).
* *
* For local kernel threads that are spawned from system calls handler when a * For local kernel threads that are spawned from system calls handler when a
@ -739,7 +761,7 @@ static const struct file_operations kcov_fops = {
* *
* See Documentation/dev-tools/kcov.rst for more details. * See Documentation/dev-tools/kcov.rst for more details.
* *
* Internally, this function looks up the kcov device associated with the * Internally, kcov_remote_start() looks up the kcov device associated with the
* provided handle, allocates an area for coverage collection, and saves the * provided handle, allocates an area for coverage collection, and saves the
* pointers to kcov and area into the current task_struct to allow coverage to * pointers to kcov and area into the current task_struct to allow coverage to
* be collected via __sanitizer_cov_trace_pc() * be collected via __sanitizer_cov_trace_pc()
@ -752,6 +774,39 @@ static inline bool kcov_mode_enabled(unsigned int mode)
return (mode & ~KCOV_IN_CTXSW) != KCOV_MODE_DISABLED; return (mode & ~KCOV_IN_CTXSW) != KCOV_MODE_DISABLED;
} }
void kcov_remote_softirq_start(struct task_struct *t)
{
struct kcov_percpu_data *data = this_cpu_ptr(&kcov_percpu_data);
unsigned int mode;
mode = READ_ONCE(t->kcov_mode);
barrier();
if (kcov_mode_enabled(mode)) {
data->saved_mode = mode;
data->saved_size = t->kcov_size;
data->saved_area = t->kcov_area;
data->saved_sequence = t->kcov_sequence;
data->saved_kcov = t->kcov;
kcov_stop(t);
}
}
void kcov_remote_softirq_stop(struct task_struct *t)
{
struct kcov_percpu_data *data = this_cpu_ptr(&kcov_percpu_data);
if (data->saved_kcov) {
kcov_start(t, data->saved_kcov, data->saved_size,
data->saved_area, data->saved_mode,
data->saved_sequence);
data->saved_mode = 0;
data->saved_size = 0;
data->saved_area = NULL;
data->saved_sequence = 0;
data->saved_kcov = NULL;
}
}
void kcov_remote_start(u64 handle) void kcov_remote_start(u64 handle)
{ {
struct task_struct *t = current; struct task_struct *t = current;
@ -761,28 +816,42 @@ void kcov_remote_start(u64 handle)
void *area; void *area;
unsigned int size; unsigned int size;
int sequence; int sequence;
unsigned long flags;
if (WARN_ON(!kcov_check_handle(handle, true, true, true))) if (WARN_ON(!kcov_check_handle(handle, true, true, true)))
return; return;
if (WARN_ON(!in_task())) if (!in_task() && !in_serving_softirq())
return; return;
local_irq_save(flags);
/* /*
* Check that kcov_remote_start is not called twice * Check that kcov_remote_start() is not called twice in background
* nor called by user tasks (with enabled kcov). * threads nor called by user tasks (with enabled kcov).
*/ */
mode = READ_ONCE(t->kcov_mode); mode = READ_ONCE(t->kcov_mode);
if (WARN_ON(kcov_mode_enabled(mode))) if (WARN_ON(in_task() && kcov_mode_enabled(mode))) {
local_irq_restore(flags);
return; return;
}
kcov_debug("handle = %llx\n", handle); /*
* Check that kcov_remote_start() is not called twice in softirqs.
* Note, that kcov_remote_start() can be called from a softirq that
* happened while collecting coverage from a background thread.
*/
if (WARN_ON(in_serving_softirq() && t->kcov_softirq)) {
local_irq_restore(flags);
return;
}
spin_lock(&kcov_remote_lock); spin_lock(&kcov_remote_lock);
remote = kcov_remote_find(handle); remote = kcov_remote_find(handle);
if (!remote) { if (!remote) {
spin_unlock(&kcov_remote_lock); spin_unlock_irqrestore(&kcov_remote_lock, flags);
return; return;
} }
kcov_debug("handle = %llx, context: %s\n", handle,
in_task() ? "task" : "softirq");
kcov = remote->kcov; kcov = remote->kcov;
/* Put in kcov_remote_stop(). */ /* Put in kcov_remote_stop(). */
kcov_get(kcov); kcov_get(kcov);
@ -790,12 +859,18 @@ void kcov_remote_start(u64 handle)
* Read kcov fields before unlock to prevent races with * Read kcov fields before unlock to prevent races with
* KCOV_DISABLE / kcov_remote_reset(). * KCOV_DISABLE / kcov_remote_reset().
*/ */
size = kcov->remote_size;
mode = kcov->mode; mode = kcov->mode;
sequence = kcov->sequence; sequence = kcov->sequence;
area = kcov_remote_area_get(size); if (in_task()) {
spin_unlock(&kcov_remote_lock); size = kcov->remote_size;
area = kcov_remote_area_get(size);
} else {
size = CONFIG_KCOV_IRQ_AREA_SIZE;
area = this_cpu_ptr(&kcov_percpu_data)->irq_area;
}
spin_unlock_irqrestore(&kcov_remote_lock, flags);
/* Can only happen when in_task(). */
if (!area) { if (!area) {
area = vmalloc(size * sizeof(unsigned long)); area = vmalloc(size * sizeof(unsigned long));
if (!area) { if (!area) {
@ -803,11 +878,20 @@ void kcov_remote_start(u64 handle)
return; return;
} }
} }
local_irq_save(flags);
/* Reset coverage size. */ /* Reset coverage size. */
*(u64 *)area = 0; *(u64 *)area = 0;
if (in_serving_softirq()) {
kcov_remote_softirq_start(t);
t->kcov_softirq = 1;
}
kcov_start(t, kcov, size, area, mode, sequence); kcov_start(t, kcov, size, area, mode, sequence);
local_irq_restore(flags);
} }
EXPORT_SYMBOL(kcov_remote_start); EXPORT_SYMBOL(kcov_remote_start);
@ -875,31 +959,53 @@ void kcov_remote_stop(void)
void *area; void *area;
unsigned int size; unsigned int size;
int sequence; int sequence;
unsigned long flags;
if (!in_task() && !in_serving_softirq())
return;
local_irq_save(flags);
mode = READ_ONCE(t->kcov_mode); mode = READ_ONCE(t->kcov_mode);
barrier(); barrier();
if (!kcov_mode_enabled(mode)) if (!kcov_mode_enabled(mode)) {
local_irq_restore(flags);
return; return;
}
kcov = t->kcov; kcov = t->kcov;
area = t->kcov_area; area = t->kcov_area;
size = t->kcov_size; size = t->kcov_size;
sequence = t->kcov_sequence; sequence = t->kcov_sequence;
if (WARN_ON(!in_serving_softirq() && t->kcov_softirq)) {
local_irq_restore(flags);
return;
}
kcov_stop(t); kcov_stop(t);
if (in_serving_softirq()) {
t->kcov_softirq = 0;
kcov_remote_softirq_stop(t);
}
spin_lock(&kcov->lock); spin_lock(&kcov->lock);
/* /*
* KCOV_DISABLE could have been called between kcov_remote_start() * KCOV_DISABLE could have been called between kcov_remote_start()
* and kcov_remote_stop(), hence the check. * and kcov_remote_stop(), hence the sequence check.
*/ */
if (sequence == kcov->sequence && kcov->remote) if (sequence == kcov->sequence && kcov->remote)
kcov_move_area(kcov->mode, kcov->area, kcov->size, area); kcov_move_area(kcov->mode, kcov->area, kcov->size, area);
spin_unlock(&kcov->lock); spin_unlock(&kcov->lock);
spin_lock(&kcov_remote_lock); if (in_task()) {
kcov_remote_area_put(area, size); spin_lock(&kcov_remote_lock);
spin_unlock(&kcov_remote_lock); kcov_remote_area_put(area, size);
spin_unlock(&kcov_remote_lock);
}
local_irq_restore(flags);
/* Get in kcov_remote_start(). */
kcov_put(kcov); kcov_put(kcov);
} }
EXPORT_SYMBOL(kcov_remote_stop); EXPORT_SYMBOL(kcov_remote_stop);
@ -913,6 +1019,16 @@ EXPORT_SYMBOL(kcov_common_handle);
static int __init kcov_init(void) static int __init kcov_init(void)
{ {
int cpu;
for_each_possible_cpu(cpu) {
void *area = vmalloc(CONFIG_KCOV_IRQ_AREA_SIZE *
sizeof(unsigned long));
if (!area)
return -ENOMEM;
per_cpu_ptr(&kcov_percpu_data, cpu)->irq_area = area;
}
/* /*
* The kcov debugfs file won't ever get removed and thus, * The kcov debugfs file won't ever get removed and thus,
* there is no need to protect it against removal races. The * there is no need to protect it against removal races. The

9
lib/Kconfig.debug
View File

@ -1774,6 +1774,15 @@ config KCOV_INSTRUMENT_ALL
filesystem fuzzing with AFL) then you will want to enable coverage filesystem fuzzing with AFL) then you will want to enable coverage
for more specific subsets of files, and should say n here. for more specific subsets of files, and should say n here.
config KCOV_IRQ_AREA_SIZE
hex "Size of interrupt coverage collection area in words"
depends on KCOV
default 0x40000
help
KCOV uses preallocated per-cpu areas to collect coverage from
soft interrupts. This specifies the size of those areas in the
number of unsigned long words.
menuconfig RUNTIME_TESTING_MENU menuconfig RUNTIME_TESTING_MENU
bool "Runtime Testing" bool "Runtime Testing"
def_bool y def_bool y