Skip to content

Commit

Permalink
profiling: attempt to remove per-cpu profile flip buffer
Browse files Browse the repository at this point in the history 
This is the really old legacy kernel profiling code, which has long
since been obviated by "real profiling" (ie 'prof' and company), and
mainly remains as a source of syzbot reports.

There are anecdotal reports that people still use it for boot-time
profiling, but it's unlikely that such use would care about the old NUMA
optimizations in this code from 2004 (commit ad02973: "profile: 512x
Altix timer interrupt livelock fix" in the BK import archive at [1])

So in order to head off future syzbot reports, let's try to simplify
this code and get rid of the per-cpu profile buffers that are quite a
large portion of the complexity footprint of this thing (including CPU
hotplug callbacks etc).

It's unlikely anybody will actually notice, or possibly, as Thomas put
it: "Only people who indulge in nostalgia will notice :)".

That said, if it turns out that this code is actually actively used by
somebody, we can always revert this removal.  Thus the "attempt" in the
summary line.

[ Note: in a small nod to "the profiling code can cause NUMA problems",
  this also removes the "increment the last entry in the profiling array
  on any unknown hits" logic. That would account any program counter in
  a module to that single counter location, and might exacerbate any
  NUMA cacheline bouncing issues ]

Link: https://lore.kernel.org/all/CAHk-=wgs52BxT4Zjmjz8aNvHWKxf5_ThBY4bYL1Y6CTaNL2dTw@mail.gmail.com/
Link:  https://git.kernel.org/pub/scm/linux/kernel/git/tglx/history.git [1]
Cc: Thomas Gleixner <[email protected]>
Cc: Tetsuo Handa <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
  • Loading branch information
@torvalds
torvalds committed Jul 29, 2024
1 parent 7c51f7b commit 2accfdb
Show file tree
Hide file tree
Showing 2 changed files with 2 additions and 182 deletions.
1 change: 0 additions & 1 deletion include/linux/cpuhotplug.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -100,7 +100,6 @@ enum cpuhp_state {
CPUHP_WORKQUEUE_PREP,
CPUHP_POWER_NUMA_PREPARE,
CPUHP_HRTIMERS_PREPARE,
CPUHP_PROFILE_PREPARE,
CPUHP_X2APIC_PREPARE,
CPUHP_SMPCFD_PREPARE,
CPUHP_RELAY_PREPARE,
Expand Down
183 changes: 2 additions & 181 deletions kernel/profile.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -129,180 +129,13 @@ int __ref profile_init(void)
return -ENOMEM;
}

#if defined(CONFIG_SMP) && defined(CONFIG_PROC_FS)
/*
* Each cpu has a pair of open-addressed hashtables for pending
* profile hits. read_profile() IPI's all cpus to request them
* to flip buffers and flushes their contents to prof_buffer itself.
* Flip requests are serialized by the profile_flip_mutex. The sole
* use of having a second hashtable is for avoiding cacheline
* contention that would otherwise happen during flushes of pending
* profile hits required for the accuracy of reported profile hits
* and so resurrect the interrupt livelock issue.
*
* The open-addressed hashtables are indexed by profile buffer slot
* and hold the number of pending hits to that profile buffer slot on
* a cpu in an entry. When the hashtable overflows, all pending hits
* are accounted to their corresponding profile buffer slots with
* atomic_add() and the hashtable emptied. As numerous pending hits
* may be accounted to a profile buffer slot in a hashtable entry,
* this amortizes a number of atomic profile buffer increments likely
* to be far larger than the number of entries in the hashtable,
* particularly given that the number of distinct profile buffer
* positions to which hits are accounted during short intervals (e.g.
* several seconds) is usually very small. Exclusion from buffer
* flipping is provided by interrupt disablement (note that for
* SCHED_PROFILING or SLEEP_PROFILING profile_hit() may be called from
* process context).
* The hash function is meant to be lightweight as opposed to strong,
* and was vaguely inspired by ppc64 firmware-supported inverted
* pagetable hash functions, but uses a full hashtable full of finite
* collision chains, not just pairs of them.
*
* -- nyc
*/
static void __profile_flip_buffers(void *unused)
{
int cpu = smp_processor_id();

per_cpu(cpu_profile_flip, cpu) = !per_cpu(cpu_profile_flip, cpu);
}

static void profile_flip_buffers(void)
{
int i, j, cpu;

mutex_lock(&profile_flip_mutex);
j = per_cpu(cpu_profile_flip, get_cpu());
put_cpu();
on_each_cpu(__profile_flip_buffers, NULL, 1);
for_each_online_cpu(cpu) {
struct profile_hit *hits = per_cpu(cpu_profile_hits, cpu)[j];
for (i = 0; i < NR_PROFILE_HIT; ++i) {
if (!hits[i].hits) {
if (hits[i].pc)
hits[i].pc = 0;
continue;
}
atomic_add(hits[i].hits, &prof_buffer[hits[i].pc]);
hits[i].hits = hits[i].pc = 0;
}
}
mutex_unlock(&profile_flip_mutex);
}

static void profile_discard_flip_buffers(void)
{
int i, cpu;

mutex_lock(&profile_flip_mutex);
i = per_cpu(cpu_profile_flip, get_cpu());
put_cpu();
on_each_cpu(__profile_flip_buffers, NULL, 1);
for_each_online_cpu(cpu) {
struct profile_hit *hits = per_cpu(cpu_profile_hits, cpu)[i];
memset(hits, 0, NR_PROFILE_HIT*sizeof(struct profile_hit));
}
mutex_unlock(&profile_flip_mutex);
}

static void do_profile_hits(int type, void *__pc, unsigned int nr_hits)
{
unsigned long primary, secondary, flags, pc = (unsigned long)__pc;
int i, j, cpu;
struct profile_hit *hits;

pc = min((pc - (unsigned long)_stext) >> prof_shift, prof_len - 1);
i = primary = (pc & (NR_PROFILE_GRP - 1)) << PROFILE_GRPSHIFT;
secondary = (~(pc << 1) & (NR_PROFILE_GRP - 1)) << PROFILE_GRPSHIFT;
cpu = get_cpu();
hits = per_cpu(cpu_profile_hits, cpu)[per_cpu(cpu_profile_flip, cpu)];
if (!hits) {
put_cpu();
return;
}
/*
* We buffer the global profiler buffer into a per-CPU
* queue and thus reduce the number of global (and possibly
* NUMA-alien) accesses. The write-queue is self-coalescing:
*/
local_irq_save(flags);
do {
for (j = 0; j < PROFILE_GRPSZ; ++j) {
if (hits[i + j].pc == pc) {
hits[i + j].hits += nr_hits;
goto out;
} else if (!hits[i + j].hits) {
hits[i + j].pc = pc;
hits[i + j].hits = nr_hits;
goto out;
}
}
i = (i + secondary) & (NR_PROFILE_HIT - 1);
} while (i != primary);

/*
* Add the current hit(s) and flush the write-queue out
* to the global buffer:
*/
atomic_add(nr_hits, &prof_buffer[pc]);
for (i = 0; i < NR_PROFILE_HIT; ++i) {
atomic_add(hits[i].hits, &prof_buffer[hits[i].pc]);
hits[i].pc = hits[i].hits = 0;
}
out:
local_irq_restore(flags);
put_cpu();
}

static int profile_dead_cpu(unsigned int cpu)
{
struct page *page;
int i;

for (i = 0; i < 2; i++) {
if (per_cpu(cpu_profile_hits, cpu)[i]) {
page = virt_to_page(per_cpu(cpu_profile_hits, cpu)[i]);
per_cpu(cpu_profile_hits, cpu)[i] = NULL;
__free_page(page);
}
}
return 0;
}

static int profile_prepare_cpu(unsigned int cpu)
{
int i, node = cpu_to_mem(cpu);
struct page *page;

per_cpu(cpu_profile_flip, cpu) = 0;

for (i = 0; i < 2; i++) {
if (per_cpu(cpu_profile_hits, cpu)[i])
continue;

page = __alloc_pages_node(node, GFP_KERNEL | __GFP_ZERO, 0);
if (!page) {
profile_dead_cpu(cpu);
return -ENOMEM;
}
per_cpu(cpu_profile_hits, cpu)[i] = page_address(page);

}
return 0;
}

#else /* !CONFIG_SMP */
#define profile_flip_buffers() do { } while (0)
#define profile_discard_flip_buffers() do { } while (0)

static void do_profile_hits(int type, void *__pc, unsigned int nr_hits)
{
unsigned long pc;
pc = ((unsigned long)__pc - (unsigned long)_stext) >> prof_shift;
atomic_add(nr_hits, &prof_buffer[min(pc, prof_len - 1)]);
if (pc < prof_len)
atomic_add(nr_hits, &prof_buffer[pc]);
}
#endif /* !CONFIG_SMP */

void profile_hits(int type, void *__pc, unsigned int nr_hits)
{
Expand Down Expand Up @@ -340,7 +173,6 @@ read_profile(struct file *file, char __user *buf, size_t count, loff_t *ppos)
char *pnt;
unsigned long sample_step = 1UL << prof_shift;

profile_flip_buffers();
if (p >= (prof_len+1)*sizeof(unsigned int))
return 0;
if (count > (prof_len+1)*sizeof(unsigned int) - p)
Expand Down Expand Up @@ -386,7 +218,6 @@ static ssize_t write_profile(struct file *file, const char __user *buf,
return -EINVAL;
}
#endif
profile_discard_flip_buffers();
memset(prof_buffer, 0, prof_len * sizeof(atomic_t));
return count;
}
Expand All @@ -404,20 +235,10 @@ int __ref create_proc_profile(void)

if (!prof_on)
return 0;
#ifdef CONFIG_SMP
err = cpuhp_setup_state(CPUHP_PROFILE_PREPARE, "PROFILE_PREPARE",
profile_prepare_cpu, profile_dead_cpu);
if (err)
return err;
#endif
entry = proc_create("profile", S_IWUSR | S_IRUGO,
NULL, &profile_proc_ops);
if (entry)
proc_set_size(entry, (1 + prof_len) * sizeof(atomic_t));
#ifdef CONFIG_SMP
else
cpuhp_remove_state(CPUHP_PROFILE_PREPARE);
#endif
return err;
}
subsys_initcall(create_proc_profile);
Expand Down

0 comments on commit 2accfdb

Please sign in to comment.