Re: [PATCH 14/16] mm/slab_common: drop kmem_alloc & avoid dereferencing fields when not using

From: Hyeonggon Yoo
Date: Tue Aug 02 2022 - 05:22:48 EST

Next message: Jiucheng Xu: "Re: [PATCH v3 4/4] dt-binding: perf: Add Amlogic DDR PMU"
Previous message: Geert Uytterhoeven: "Re: Build regressions/improvements in v5.19"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On Fri, Jul 29, 2022 at 01:23:38PM +0200, Vlastimil Babka wrote:
> On 7/12/22 15:39, Hyeonggon Yoo wrote:
> > Drop kmem_alloc event class, and define kmalloc and kmem_cache_alloc
> > using TRACE_EVENT() macro.
> >
> > And then this patch does:
> > - Do not pass pointer to struct kmem_cache to trace_kmalloc.
> > gfp flag is enough to know if it's accounted or not.
> > - Avoid dereferencing s->object_size and s->size when not using kmem_cache_alloc event.
> > - Avoid dereferencing s->name in when not using kmem_cache_free event.
> >
> > Suggested-by: Vlastimil Babka <vbabka@xxxxxxx>
> > Signed-off-by: Hyeonggon Yoo <42.hyeyoo@xxxxxxxxx>
>
> Reviewed-by: Vlastimil Babka <vbabka@xxxxxxx>
>
> With some comments:
>
> > ---
> > include/trace/events/kmem.h | 64 +++++++++++++++++++++++++------------
> > mm/slab.c | 16 ++++------
> > mm/slab_common.c | 17 ++++++----
> > mm/slob.c | 16 +++-------
> > mm/slub.c | 13 +++-----
> > 5 files changed, 70 insertions(+), 56 deletions(-)
> >
> > diff --git a/include/trace/events/kmem.h b/include/trace/events/kmem.h
> > index e078ebcdc4b1..0ded2c351062 100644
> > --- a/include/trace/events/kmem.h
> > +++ b/include/trace/events/kmem.h
> > @@ -9,17 +9,15 @@
> > #include <linux/tracepoint.h>
> > #include <trace/events/mmflags.h>
> >
> > -DECLARE_EVENT_CLASS(kmem_alloc,
> > +TRACE_EVENT(kmem_cache_alloc,
> >
> > TP_PROTO(unsigned long call_site,
> > const void *ptr,
> > struct kmem_cache *s,
> > - size_t bytes_req,
> > - size_t bytes_alloc,
> > gfp_t gfp_flags,
> > int node),
> >
> > - TP_ARGS(call_site, ptr, s, bytes_req, bytes_alloc, gfp_flags, node),
> > + TP_ARGS(call_site, ptr, s, gfp_flags, node),
> >
> > TP_STRUCT__entry(
> > __field( unsigned long, call_site )
> > @@ -34,8 +32,8 @@ DECLARE_EVENT_CLASS(kmem_alloc,
> > TP_fast_assign(
> > __entry->call_site = call_site;
> > __entry->ptr = ptr;
> > - __entry->bytes_req = bytes_req;
> > - __entry->bytes_alloc = bytes_alloc;
> > + __entry->bytes_req = s->object_size;
> > + __entry->bytes_alloc = s->size;
> > __entry->gfp_flags = (__force unsigned long)gfp_flags;
> > __entry->node = node;
> > __entry->accounted = IS_ENABLED(CONFIG_MEMCG_KMEM) ?
>
> This assignment continues by ...
> (s && s->flags & SLAB_ACCOUNT)) : false;
>
> We could now remove the "s &&" as we assume it's non-NULL anyway.

Right. will adjust.

>
> > @@ -53,22 +51,46 @@ DECLARE_EVENT_CLASS(kmem_alloc,
> > __entry->accounted ? "true" : "false")
> > );
> >
> > -DEFINE_EVENT(kmem_alloc, kmalloc,
> > +TRACE_EVENT(kmalloc,
> >
> > - TP_PROTO(unsigned long call_site, const void *ptr,
> > - struct kmem_cache *s, size_t bytes_req, size_t bytes_alloc,
> > - gfp_t gfp_flags, int node),
> > + TP_PROTO(unsigned long call_site,
> > + const void *ptr,
> > + size_t bytes_req,
> > + size_t bytes_alloc,
> > + gfp_t gfp_flags,
> > + int node),
> >
> > - TP_ARGS(call_site, ptr, s, bytes_req, bytes_alloc, gfp_flags, node)
> > -);
> > + TP_ARGS(call_site, ptr, bytes_req, bytes_alloc, gfp_flags, node),
> >
> > -DEFINE_EVENT(kmem_alloc, kmem_cache_alloc,
> > + TP_STRUCT__entry(
> > + __field( unsigned long, call_site )
> > + __field( const void *, ptr )
> > + __field( size_t, bytes_req )
> > + __field( size_t, bytes_alloc )
> > + __field( unsigned long, gfp_flags )
> > + __field( int, node )
> > + __field( bool, accounted )
> > + ),
> >
> > - TP_PROTO(unsigned long call_site, const void *ptr,
> > - struct kmem_cache *s, size_t bytes_req, size_t bytes_alloc,
> > - gfp_t gfp_flags, int node),
> > + TP_fast_assign(
> > + __entry->call_site = call_site;
> > + __entry->ptr = ptr;
> > + __entry->bytes_req = bytes_req;
> > + __entry->bytes_alloc = bytes_alloc;
> > + __entry->gfp_flags = (__force unsigned long)gfp_flags;
> > + __entry->node = node;
> > + __entry->accounted = IS_ENABLED(CONFIG_MEMCG_KMEM) ?
> > + (gfp_flags & __GFP_ACCOUNT) : false;
>
> This test could be perhaps moved to TP_printk and the accounted field
> removed completely, but we need to make sure trace-cmd (which reads the
> binary entries and does the printk in userspace) can interpret the tests
> properly, including the IS_ENABLED(CONFIG_MEMCG_KMEM) part.

Okay, I'll try moving it and check if trace-cmd could understand it.

>
> > + ),
> >
> > - TP_ARGS(call_site, ptr, s, bytes_req, bytes_alloc, gfp_flags, node)
> > + TP_printk("call_site=%pS ptr=%p bytes_req=%zu bytes_alloc=%zu gfp_flags=%s node=%d accounted=%s",
> > + (void *)__entry->call_site,
> > + __entry->ptr,
> > + __entry->bytes_req,
> > + __entry->bytes_alloc,
> > + show_gfp_flags(__entry->gfp_flags),
> > + __entry->node,
> > + __entry->accounted ? "true" : "false")
> > );
> >
> > TRACE_EVENT(kfree,
> > @@ -93,20 +115,20 @@ TRACE_EVENT(kfree,
> >
> > TRACE_EVENT(kmem_cache_free,
> >
> > - TP_PROTO(unsigned long call_site, const void *ptr, const char *name),
> > + TP_PROTO(unsigned long call_site, const void *ptr, const struct kmem_cache *s),
> >
> > - TP_ARGS(call_site, ptr, name),
> > + TP_ARGS(call_site, ptr, s),
> >
> > TP_STRUCT__entry(
> > __field( unsigned long, call_site )
> > __field( const void *, ptr )
> > - __string( name, name )
> > + __string( name, s->name )
> > ),
> >
> > TP_fast_assign(
> > __entry->call_site = call_site;
> > __entry->ptr = ptr;
> > - __assign_str(name, name);
> > + __assign_str(name, s->name);
> > ),
> >
> > TP_printk("call_site=%pS ptr=%p name=%s",
> > diff --git a/mm/slab.c b/mm/slab.c
> > index f9b74831e3f4..1685e5507a59 100644
> > --- a/mm/slab.c
> > +++ b/mm/slab.c
> > @@ -3447,8 +3447,7 @@ void *__kmem_cache_alloc_lru(struct kmem_cache *cachep, struct list_lru *lru,
> > {
> > void *ret = slab_alloc(cachep, lru, flags, cachep->object_size, _RET_IP_);
> >
> > - trace_kmem_cache_alloc(_RET_IP_, ret, cachep, cachep->object_size,
> > - cachep->size, flags, NUMA_NO_NODE);
> > + trace_kmem_cache_alloc(_RET_IP_, ret, cachep, flags, NUMA_NO_NODE);
> >
> > return ret;
> > }
> > @@ -3537,8 +3536,9 @@ kmem_cache_alloc_trace(struct kmem_cache *cachep, gfp_t flags, size_t size)
> > ret = slab_alloc(cachep, NULL, flags, size, _RET_IP_);
> >
> > ret = kasan_kmalloc(cachep, ret, size, flags);
> > - trace_kmalloc(_RET_IP_, ret, cachep,
> > - size, cachep->size, flags, NUMA_NO_NODE);
> > + trace_kmalloc(_RET_IP_, ret,
> > + size, cachep->size,
> > + flags, NUMA_NO_NODE);
>
> This could now fit on 2, maybe even 1 line?
> Other trace_* calls could be perhaps also reduced, please check.
>

Okay, will check.

> > return ret;
> > }
> > EXPORT_SYMBOL(kmem_cache_alloc_trace);
> > @@ -3561,9 +3561,7 @@ void *kmem_cache_alloc_node(struct kmem_cache *cachep, gfp_t flags, int nodeid)
> > {
> > void *ret = slab_alloc_node(cachep, NULL, flags, nodeid, cachep->object_size, _RET_IP_);
> >
> > - trace_kmem_cache_alloc(_RET_IP_, ret, cachep,
> > - cachep->object_size, cachep->size,
> > - flags, nodeid);
> > + trace_kmem_cache_alloc(_RET_IP_, ret, cachep, flags, nodeid);
> >
> > return ret;
> > }
> > @@ -3588,7 +3586,7 @@ void *kmem_cache_alloc_node_trace(struct kmem_cache *cachep,
> > ret = slab_alloc_node(cachep, NULL, flags, nodeid, size, _RET_IP_);
> >
> > ret = kasan_kmalloc(cachep, ret, size, flags);
> > - trace_kmalloc(_RET_IP_, ret, cachep,
> > + trace_kmalloc(_RET_IP_, ret,
> > size, cachep->size,
> > flags, nodeid);
> > return ret;
> > @@ -3652,7 +3650,7 @@ void kmem_cache_free(struct kmem_cache *cachep, void *objp)
> > if (!cachep)
> > return;
> >
> > - trace_kmem_cache_free(_RET_IP_, objp, cachep->name);
> > + trace_kmem_cache_free(_RET_IP_, objp, cachep);
> > __do_kmem_cache_free(cachep, objp, _RET_IP_);
> > }
> > EXPORT_SYMBOL(kmem_cache_free);
> > diff --git a/mm/slab_common.c b/mm/slab_common.c
> > index 0e66b4911ebf..c01c6b8f0d34 100644
> > --- a/mm/slab_common.c
> > +++ b/mm/slab_common.c
> > @@ -933,7 +933,7 @@ void *__do_kmalloc_node(size_t size, gfp_t flags, int node, unsigned long caller
> >
> > if (unlikely(size > KMALLOC_MAX_CACHE_SIZE)) {
> > ret = kmalloc_large_node_notrace(size, flags, node);
> > - trace_kmalloc(_RET_IP_, ret, NULL,
> > + trace_kmalloc(_RET_IP_, ret,
> > size, PAGE_SIZE << get_order(size),
> > flags, node);
> > return ret;
> > @@ -946,8 +946,9 @@ void *__do_kmalloc_node(size_t size, gfp_t flags, int node, unsigned long caller
> >
> > ret = __kmem_cache_alloc_node(s, flags, node, size, caller);
> > ret = kasan_kmalloc(s, ret, size, flags);
> > - trace_kmalloc(_RET_IP_, ret, s, size,
> > - s->size, flags, node);
> > + trace_kmalloc(_RET_IP_, ret,
> > + size, s->size,
> > + flags, node);
> > return ret;
> > }
> >
> > @@ -1075,8 +1076,9 @@ void *kmalloc_large(size_t size, gfp_t flags)
> > {
> > void *ret = __kmalloc_large_node_notrace(size, flags, NUMA_NO_NODE);
> >
> > - trace_kmalloc(_RET_IP_, ret, NULL, size,
> > - PAGE_SIZE << get_order(size), flags, NUMA_NO_NODE);
> > + trace_kmalloc(_RET_IP_, ret,
> > + size, PAGE_SIZE << get_order(size),
> > + flags, NUMA_NO_NODE);
> > return ret;
> > }
> > EXPORT_SYMBOL(kmalloc_large);
> > @@ -1090,8 +1092,9 @@ void *kmalloc_large_node(size_t size, gfp_t flags, int node)
> > {
> > void *ret = __kmalloc_large_node_notrace(size, flags, node);
> >
> > - trace_kmalloc(_RET_IP_, ret, NULL, size,
> > - PAGE_SIZE << get_order(size), flags, node);
> > + trace_kmalloc(_RET_IP_, ret,
> > + size, PAGE_SIZE << get_order(size),
> > + flags, node);
> > return ret;
> > }
> > EXPORT_SYMBOL(kmalloc_large_node);
> > diff --git a/mm/slob.c b/mm/slob.c
> > index a4d50d957c25..97a4d2407f96 100644
> > --- a/mm/slob.c
> > +++ b/mm/slob.c
> > @@ -507,8 +507,7 @@ __do_kmalloc_node(size_t size, gfp_t gfp, int node, unsigned long caller)
> > *m = size;
> > ret = (void *)m + minalign;
> >
> > - trace_kmalloc(caller, ret, NULL,
> > - size, size + minalign, gfp, node);
> > + trace_kmalloc(caller, ret, size, size + minalign, gfp, node);
> > } else {
> > unsigned int order = get_order(size);
> >
> > @@ -516,8 +515,7 @@ __do_kmalloc_node(size_t size, gfp_t gfp, int node, unsigned long caller)
> > gfp |= __GFP_COMP;
> > ret = slob_new_pages(gfp, order, node);
> >
> > - trace_kmalloc(caller, ret, NULL,
> > - size, PAGE_SIZE << order, gfp, node);
> > + trace_kmalloc(caller, ret, size, PAGE_SIZE << order, gfp, node);
> > }
> >
> > kmemleak_alloc(ret, size, 1, gfp);
> > @@ -608,14 +606,10 @@ static void *slob_alloc_node(struct kmem_cache *c, gfp_t flags, int node)
> >
> > if (c->size < PAGE_SIZE) {
> > b = slob_alloc(c->size, flags, c->align, node, 0);
> > - trace_kmem_cache_alloc(_RET_IP_, b, NULL, c->object_size,
> > - SLOB_UNITS(c->size) * SLOB_UNIT,
>
> Hm we lose the SLOB_UNITS() adjustment here.

Thank you for catching this.

> Maybe we don't care that much,
> maybe simply c->size could be already stored with SLOB_UNITS() (but didn't
> check if other stuff breaks with that).

Hmm will try that way!

Thanks!

>
> Thanks!
>
> > - flags, node);
> > + trace_kmem_cache_alloc(_RET_IP_, b, c, flags, node);
> > } else {
> > b = slob_new_pages(flags, get_order(c->size), node);
> > - trace_kmem_cache_alloc(_RET_IP_, b, NULL, c->object_size,
> > - PAGE_SIZE << get_order(c->size),
> > - flags, node);
> > + trace_kmem_cache_alloc(_RET_IP_, b, c, flags, node);
> > }
> >
> > if (b && c->ctor) {
> > @@ -671,7 +665,7 @@ static void kmem_rcu_free(struct rcu_head *head)
> > void kmem_cache_free(struct kmem_cache *c, void *b)
> > {
> > kmemleak_free_recursive(b, c->flags);
> > - trace_kmem_cache_free(_RET_IP_, b, c->name);
> > + trace_kmem_cache_free(_RET_IP_, b, c);
> > if (unlikely(c->flags & SLAB_TYPESAFE_BY_RCU)) {
> > struct slob_rcu *slob_rcu;
> > slob_rcu = b + (c->size - sizeof(struct slob_rcu));
> > diff --git a/mm/slub.c b/mm/slub.c
> > index f1aa51480dc4..a49c69469c64 100644
> > --- a/mm/slub.c
> > +++ b/mm/slub.c
> > @@ -3243,8 +3243,7 @@ void *__kmem_cache_alloc_lru(struct kmem_cache *s, struct list_lru *lru,
> > {
> > void *ret = slab_alloc(s, lru, gfpflags, _RET_IP_, s->object_size);
> >
> > - trace_kmem_cache_alloc(_RET_IP_, ret, s, s->object_size,
> > - s->size, gfpflags, NUMA_NO_NODE);
> > + trace_kmem_cache_alloc(_RET_IP_, ret, s, gfpflags, NUMA_NO_NODE);
> >
> > return ret;
> > }
> > @@ -3274,7 +3273,7 @@ void *__kmem_cache_alloc_node(struct kmem_cache *s, gfp_t gfpflags,
> > void *kmem_cache_alloc_trace(struct kmem_cache *s, gfp_t gfpflags, size_t size)
> > {
> > void *ret = slab_alloc(s, NULL, gfpflags, _RET_IP_, size);
> > - trace_kmalloc(_RET_IP_, ret, s, size, s->size, gfpflags, NUMA_NO_NODE);
> > + trace_kmalloc(_RET_IP_, ret, size, s->size, gfpflags, NUMA_NO_NODE);
> > ret = kasan_kmalloc(s, ret, size, gfpflags);
> > return ret;
> > }
> > @@ -3285,8 +3284,7 @@ void *kmem_cache_alloc_node(struct kmem_cache *s, gfp_t gfpflags, int node)
> > {
> > void *ret = slab_alloc_node(s, NULL, gfpflags, node, _RET_IP_, s->object_size);
> >
> > - trace_kmem_cache_alloc(_RET_IP_, ret, s,
> > - s->object_size, s->size, gfpflags, node);
> > + trace_kmem_cache_alloc(_RET_IP_, ret, s, gfpflags, node);
> >
> > return ret;
> > }
> > @@ -3299,8 +3297,7 @@ void *kmem_cache_alloc_node_trace(struct kmem_cache *s,
> > {
> > void *ret = slab_alloc_node(s, NULL, gfpflags, node, _RET_IP_, size);
> >
> > - trace_kmalloc(_RET_IP_, ret, s,
> > - size, s->size, gfpflags, node);
> > + trace_kmalloc(_RET_IP_, ret, size, s->size, gfpflags, node);
> >
> > ret = kasan_kmalloc(s, ret, size, gfpflags);
> > return ret;
> > @@ -3544,7 +3541,7 @@ void kmem_cache_free(struct kmem_cache *s, void *x)
> > s = cache_from_obj(s, x);
> > if (!s)
> > return;
> > - trace_kmem_cache_free(_RET_IP_, x, s->name);
> > + trace_kmem_cache_free(_RET_IP_, x, s);
> > slab_free(s, virt_to_slab(x), x, NULL, &x, 1, _RET_IP_);
> > }
> > EXPORT_SYMBOL(kmem_cache_free);
>

--
Thanks,
Hyeonggon

Next message: Jiucheng Xu: "Re: [PATCH v3 4/4] dt-binding: perf: Add Amlogic DDR PMU"
Previous message: Geert Uytterhoeven: "Re: Build regressions/improvements in v5.19"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]