Message-ID: <Yh0+Jld+sDy1Hkdb@ip-172-31-19-208.ap-northeast-1.compute.internal>
Date:   Mon, 28 Feb 2022 21:27:02 +0000
From:   Hyeonggon Yoo <42.hyeyoo@...il.com>
To:     Vlastimil Babka <vbabka@...e.cz>
Cc:     David Rientjes <rientjes@...gle.com>,
        Christoph Lameter <cl@...ux.com>,
        Joonsoo Kim <iamjoonsoo.kim@....com>,
        Pekka Enberg <penberg@...nel.org>,
        Roman Gushchin <roman.gushchin@...ux.dev>,
        Andrew Morton <akpm@...ux-foundation.org>, linux-mm@...ck.org,
        patches@...ts.linux.dev, linux-kernel@...r.kernel.org,
        Oliver Glitta <glittao@...il.com>,
        Faiyaz Mohammed <faiyazm@...eaurora.org>,
        Jonathan Corbet <corbet@....net>,
        Randy Dunlap <rdunlap@...radead.org>,
        Marco Elver <elver@...gle.com>,
        Karolina Drobnik <karolinadrobnik@...il.com>,
        Mike Rapoport <rppt@...ux.ibm.com>
Subject: Re: [PATCH 0/5] SLUB debugfs improvements based on stackdepot

On Mon, Feb 28, 2022 at 08:10:18PM +0100, Vlastimil Babka wrote:
> On 2/26/22 08:19, Hyeonggon Yoo wrote:
> > On Fri, Feb 25, 2022 at 07:03:13PM +0100, Vlastimil Babka wrote:
> >> Hi,
> >> 
> >> this series combines and revives patches from Oliver's last year
> >> bachelor thesis (where I was the advisor) that make SLUB's debugfs
> >> files alloc_traces and free_traces more useful.
> >> The resubmission was blocked on stackdepot changes that are now merged,
> >> as explained in patch 2.
> >> 
> > 
> > Hello. I just started review/testing this series.
> > 
> > it crashed on my system (arm64)
> 
> Hmm, interesting. On x86_64 this works for me and stackdepot is allocated
> from memblock. arm64 must have memblock freeing happen earlier or something.
> (CCing memblock experts)
> 
> > I ran with boot parameter slub_debug=U, and without KASAN.
> > So CONFIG_STACKDEPOT_ALWAYS_INIT=n.
> > 
> > void * __init memblock_alloc_try_nid(
> >                         phys_addr_t size, phys_addr_t align,
> >                         phys_addr_t min_addr, phys_addr_t max_addr,
> >                         int nid)
> > {
> >         void *ptr;
> > 
> >         memblock_dbg("%s: %llu bytes align=0x%llx nid=%d from=%pa max_addr=%pa %pS\n",
> >                      __func__, (u64)size, (u64)align, nid, &min_addr,
> >                      &max_addr, (void *)_RET_IP_);
> >         ptr = memblock_alloc_internal(size, align,
> >                                            min_addr, max_addr, nid, false);
> >         if (ptr)
> >                 memset(ptr, 0, size); <--- Crash Here
> > 
> >         return ptr;
> > }
> > 
> > It crashed during create_boot_cache() -> stack_depot_init() ->
> > memblock_alloc().
> > 
> > I think That's because, in kmem_cache_init(), both slab and memblock is not
> > available. (AFAIU memblock is not available after mem_init() because of
> > memblock_free_all(), right?)
> 
> Hm yes I see, even in x86_64 version mem_init() calls memblock_free_all().
> But then, I would expect stack_depot_init() to detect that memblock_alloc()
> returns NULL, we print ""Stack Depot hash table allocation failed,
> disabling" and disable it. Instead it seems memblock_alloc() returns
> something that's already potentially used by somebody else? Sounds like a bug?


By the way, I fixed this by allowing stack_depot_init() to be called in
kmem_cache_init() too [1] and Marco suggested that calling
stack_depot_init() depending on slub_debug parameter for simplicity. [2]

I would prefer [2], Would you take a look?

[1] https://lkml.org/lkml/2022/2/27/31

[2] https://lkml.org/lkml/2022/2/28/717

> > Thanks!
> > 
> > /*
> >  * Set up kernel memory allocators
> >  */
> > static void __init mm_init(void)
> > {
> >         /*
> >          * page_ext requires contiguous pages,
> >          * bigger than MAX_ORDER unless SPARSEMEM.
> >          */
> >         page_ext_init_flatmem();
> >         init_mem_debugging_and_hardening();
> >         kfence_alloc_pool();
> >         report_meminit();
> >         stack_depot_early_init();
> >         mem_init();
> >         mem_init_print_info();
> >         kmem_cache_init();
> >         /*
> >          * page_owner must be initialized after buddy is ready, and also after
> >          * slab is ready so that stack_depot_init() works properly
> >          */)
> > 
> >> Patch 1 is a new preparatory cleanup.
> >> 
> >> Patch 2 originally submitted here [1], was merged to mainline but
> >> reverted for stackdepot related issues as explained in the patch.
> >> 
> >> Patches 3-5 originally submitted as RFC here [2]. In this submission I
> >> have omitted the new file 'all_objects' (patch 3/3 in [2]) as it might
> >> be considered too intrusive so I will postpone it for later. The docs
> >> patch is adjusted accordingly.
> >> 
> >> Also available in git, based on v5.17-rc1:
> >> https://git.kernel.org/pub/scm/linux/kernel/git/vbabka/linux.git/log/?h=slub-stackdepot-v1
> >> 
> >> I'd like to ask for some review before I add this to the slab tree.
> >> 
> >> [1] https://lore.kernel.org/all/20210414163434.4376-1-glittao@gmail.com/
> >> [2] https://lore.kernel.org/all/20210521121127.24653-1-glittao@gmail.com/
> >> 
> >> Oliver Glitta (4):
> >>   mm/slub: use stackdepot to save stack trace in objects
> >>   mm/slub: aggregate and print stack traces in debugfs files
> >>   mm/slub: sort debugfs output by frequency of stack traces
> >>   slab, documentation: add description of debugfs files for SLUB caches
> >> 
> >> Vlastimil Babka (1):
> >>   mm/slub: move struct track init out of set_track()
> >> 
> >>  Documentation/vm/slub.rst |  61 +++++++++++++++
> >>  init/Kconfig              |   1 +
> >>  mm/slub.c                 | 152 +++++++++++++++++++++++++-------------
> >>  3 files changed, 162 insertions(+), 52 deletions(-)
> >> 
> >> -- 
> >> 2.35.1
> >> 
> >> 
> > 
> 

-- 
Thank you, You are awesome!
Hyeonggon :-)

Powered by blists - more mailing lists