Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Thu, 1 Sep 2022 20:14:40 +0900
From:   Hyeonggon Yoo <42.hyeyoo@gmail.com>
To:     Feng Tang <feng.tang@intel.com>
Cc:     Andrew Morton <akpm@linux-foundation.org>,
        Vlastimil Babka <vbabka@suse.cz>,
        Christoph Lameter <cl@linux.com>,
        Pekka Enberg <penberg@kernel.org>,
        David Rientjes <rientjes@google.com>,
        Joonsoo Kim <iamjoonsoo.kim@lge.com>,
        Roman Gushchin <roman.gushchin@linux.dev>,
        Dmitry Vyukov <dvyukov@google.com>,
        "Hansen, Dave" <dave.hansen@intel.com>,
        "linux-mm@kvack.org" <linux-mm@kvack.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        Robin Murphy <robin.murphy@arm.com>,
        John Garry <john.garry@huawei.com>,
        Kefeng Wang <wangkefeng.wang@huawei.com>
Subject: Re: [PATCH v4 1/4] mm/slub: enable debugging memory wasting of
 kmalloc
Message-ID: <YxCUIM4BWVZD6fnk@hyeyoo>
References: <20220829075618.69069-1-feng.tang@intel.com>
 <20220829075618.69069-2-feng.tang@intel.com>
 <Yw91n/AzjmAXmdh8@hyeyoo>
 <YxA9etTqGsaogfZ6@feng-clx>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <YxA9etTqGsaogfZ6@feng-clx>
Precedence: bulk

On Thu, Sep 01, 2022 at 01:04:58PM +0800, Feng Tang wrote:
> On Wed, Aug 31, 2022 at 10:52:15PM +0800, Hyeonggon Yoo wrote:
> > On Mon, Aug 29, 2022 at 03:56:15PM +0800, Feng Tang wrote:
> > > kmalloc's API family is critical for mm, with one nature that it will
> > > round up the request size to a fixed one (mostly power of 2). Say
> > > when user requests memory for '2^n + 1' bytes, actually 2^(n+1) bytes
> > > could be allocated, so in worst case, there is around 50% memory
> > > space waste.
> > > 
> > 
> > [...]
> > 
> > >  static void *___slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node,
> > > -			  unsigned long addr, struct kmem_cache_cpu *c)
> > > +			  unsigned long addr, struct kmem_cache_cpu *c, unsigned int orig_size)
> > >  {
> > >  	void *freelist;
> > >  	struct slab *slab;
> > > @@ -3115,6 +3158,7 @@ static void *___slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node,
> > >  
> > >  		if (s->flags & SLAB_STORE_USER)
> > >  			set_track(s, freelist, TRACK_ALLOC, addr);
> > > +		set_orig_size(s, freelist, orig_size);
> > >  
> > >  		return freelist;
> > >  	}
> > > @@ -3140,6 +3184,8 @@ static void *___slab_alloc(struct kmem_cache *s, gfp_t gfpflags, int node,
> > >  		 */
> > >  		if (s->flags & SLAB_STORE_USER)
> > >  			set_track(s, freelist, TRACK_ALLOC, addr);
> > > +		set_orig_size(s, freelist, orig_size);
> > > +
> > >  		return freelist;
> > >  	}
> > 
> > Maybe we can move set_track() and set_orig_size() to after slab_post_alloc_hook().
> > something like alloc/free hooks for debugging caches? (and drop orig_size parameter.) 
> 
> Yep, we discussed this during v3 review
> https://lore.kernel.org/lkml/442d2b9c-9f07-8954-b90e-b4a9f8b64303@intel.com/

Ah, I missed that :) Thanks!

Considering the added cost (should be low) and races with validation,
I think this approach will cost more than it get. Sorry for the noise.

p.s. I think I can review this series in few days.
Thanks for your efforts!

> Will revisit this considering recent refactoring and the following
> kmalloc data redzone patches.

> Thanks,
> Feng
> 
> > Thanks!

-- 
Thanks,
Hyeonggon