Date:	Sat, 16 Feb 2008 11:26:18 -0800
From:	Andrew Morton <akpm@...ux-foundation.org>
To:	Eric Dumazet <dada1@...mosbay.com>
Cc:	Herbert Xu <herbert@...dor.apana.org.au>,
	"David S. Miller" <davem@...emloft.net>, netdev@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: Re: include/linux/pcounter.h

On Sat, 16 Feb 2008 13:03:54 +0100 Eric Dumazet <dada1@...mosbay.com> wrote:

> Andrew Morton a écrit :
> > On Sat, 16 Feb 2008 11:07:29 +0100 Eric Dumazet <dada1@...mosbay.com> wrote:
> > 
> >> Andrew, pcounter is a temporary abstraction.
> > 
> > It's buggy!  Main problems are a) possible return of negative numbers b)
> > some of the API can't be from preemptible code c) excessive interrupt-off
> > time on some machines if used from irq-disabled sections.

This is starting to get tiresome.  Please do not waste my time and everyone
else's by defending the indefensible.

> a) We dont care of possibly off values when reading /proc/net/sockstat

So take the damn thing out of ./lib/ and hide it down in networking
somewhere.

>     Same arguments apply for percpu_counters.

No it does not, Not at all.  Callers regularly use percpu_counters to count
the number of some object.  These counts never go negative!

And how the heck can you say that we don't care if /proc/net/sockstat goes
negative?  You don't know that.  Someone's system-monitoring app might very
well take drastic action if it sees connection-related statistics go
negative, or around 4 gigaconnections.

> b) It is called from network parts where preemption is disabled.

It's in ./lib/

> net/ipv4/inet_timewait_sock.c:94: 
> sock_prot_inuse_add(sk->sk_prot, -1);
> net/ipv4/inet_hashtables.c:291: sock_prot_inuse_add(sk->sk_prot, 1);
> net/ipv4/inet_hashtables.c:335: sock_prot_inuse_add(sk->sk_prot, 1);
> net/ipv4/inet_hashtables.c:357: sock_prot_inuse_add(sk->sk_prot, 1);
> net/ipv4/inet_hashtables.c:390:         sock_prot_inuse_add(sk->sk_prot, -1);
> net/ipv4/raw.c:95:      sock_prot_inuse_add(sk->sk_prot, 1);
> net/ipv4/raw.c:104:             sock_prot_inuse_add(sk->sk_prot, -1);
> net/ipv4/udp.c:235:             sock_prot_inuse_add(sk->sk_prot, 1);
> net/ipv6/ipv6_sockglue.c:271: 
> sock_prot_inuse_add(sk->sk_prot, -1);
> net/ipv6/ipv6_sockglue.c:272: 
> sock_prot_inuse_add(&tcp_prot, 1);
> net/ipv6/ipv6_sockglue.c:285: 
> sock_prot_inuse_add(sk->sk_prot, -1);
> net/ipv6/ipv6_sockglue.c:286: 
> sock_prot_inuse_add(prot, 1);
> net/ipv6/inet6_hashtables.c:46: sock_prot_inuse_add(sk->sk_prot, 1);
> net/ipv6/inet6_hashtables.c:207:        sock_prot_inuse_add(sk->sk_prot, 1);
> 
> c) No need to play with irq games here.

It's in ./lib/

> > 
> >> It is temporaty because it will vanish as soon as Christoph Clameter (or 
> >> somebody else) provides real cheap per cpu counter implementation.
> > 
> > numbers?
> > 
> > most of percpu_counter_add() is only executed once per FBC_BATCH calls.
> 
> 
> 
> > 
> >> At time I introduced it in network tree (locally, not meant to invade kernel 
> >> land and makes you unhappy :) ), the goals were :
> > 
> > Well maybe as a temporary networking-only thing OK, based upon
> > performance-tested results.  But I don't think the present code is suitable
> > as part of the kernel-wide toolkit.
> > 
> >> Some counters (total sockets count) were a single integer, that were doing 
> >> ping-pong between cpus (SMP/NUMA). As they are basically lazy values (as we 
> >> dont really need to read their value), using plain atomic_t was overkill.
> >>
> >> Using a plain percpu_counters was expensive (NR_CPUS*(32+sizeof(void *)) 
> >> instead of num_possible_cpus()*4).
> > 
> > No, percpu_counters use alloc_percpu(), which is O(num_possible_cpus), not
> > O(NR_CPUS).
> 
> You are playing with words.

I assumed the comment was comparing with percpu_counters.

Seems that it was comparing with some hand-rolled static array of NR_CPUS
entries or something.

> > 
> >>  We want really fast write side. Real fast.
> > 
> > eh?  It's called on a per-connection basis, not on a per-packet basis?
> 
> Yes, per connection basis. Some workloads want to open/close more than 1000 
> sockets per second.

ie: slowpath

> You are right we also need to reduce all atomic inc/dec done per packet :)
> 
> A pcounter/perc_cpu for the device refcounter would be a very nice win, but as 
> number of devices in the system might be very big, David said no to a percpu 
> conversion. We will reconsider this when new percpu stuff can go in, so that 
> the memory cost will be minimal (4 bytes per cpu per device)
> 
> > 
> >> Read side is when you do a "cat /proc/net/sockstat".
> >> That is ... once in a while...
> > 
> > For the current single caller.  But it's in ./lib/.
> > 
> > And there's always someone out there who does whatever we don't expect them
> > to do.
> > 
> >> Now when we allocate a new socket, code to increment the "socket count" is :
> >>
> >> c03a74a8 <tcp_pcounter_add>:
> >> c03a74a8:   b8 90 26 5f c0          mov    $0xc05f2690,%eax
> >> c03a74ad:   64 8b 0d 10 f1 5e c0    mov    %fs:0xc05ef110,%ecx
> >> c03a74b4:   01 14 01                add    %edx,(%ecx,%eax,1)
> >> c03a74b7:   c3                      ret
> > 
> > I can't find that code.  I suspect that's the DEFINE_PER_CPU flavour, which
> > isn't used anywhere afaict.  Plus this omits the local_irq_save/restore (or
> > preempt_disable/enable) and the indirect function call, which can be
> > expensive.
> 
> Please do : nm vmlinux | grep tcp_pcounter_add
> 
> c03a74a8 t tcp_pcounter_add
> 
> It should be there, even if its a static function
> 

Probably - the macros hide it well.  And the function itself doesn't tell
the whole story: callers must still do preempt_dsable or local_irq_enable
and must wear the cost of an indirect call.

> > 
> >> It is even clearly stated at the top of include/linux/pcounter.h
> >>
> >> /*
> >>   * Using a dynamic percpu 'int' variable has a cost :
> >>   * 1) Extra dereference
> >>   * Current per_cpu_ptr() implementation uses an array per 'percpu variable'.
> >>   * 2) memory cost of NR_CPUS*(32+sizeof(void *)) instead of num_possible_cpus()*4
> >>   *
> >>   * This pcounter implementation is an abstraction to be able to use
> >>   * either a static or a dynamic per cpu variable.
> >>   * One dynamic per cpu variable gets a fast & cheap implementation, we can
> >>   * change pcounter implementation too.
> >>   */
> >>
> >>
> >> We all agree.
> >>
> > 
> > No we don't.  That comment is afaict wrong about the memory consumption and
> > the abstraction *isn't useful*.
> 
> Fact is that we need percpu 32bits counters, and we need to have pointers to them.
> 
> Current percpu_counters cannot cope that.

yes they can.

> > 
> > Why do we want some abstraction which makes alloc_percpu() storage and
> > DEFINE_PERCPU storage "look the same"?  What use is there in that?  One is
> > per-object storage and one is singleton storage - they're quite different
> > things and they are used in quite different situations and they are
> > basically never interchangeable.  Yet we add this pretend-they're-the-same
> > wrapper around them which costs us an indirect function call on the
> > fastpath.
> 
> I believe all 'pcounter' are in fact statically allocated 'one per struct 
> proto' to track inuse count. (search for DEFINE_PROTO_INUSE() uses)
> 
> # find net include|xargs grep -n DEFINE_PROTO_INUSE
> net/dccp/ipv6.c:1105:DEFINE_PROTO_INUSE(dccp_v6)
> net/dccp/ipv4.c:920:DEFINE_PROTO_INUSE(dccp_v4)
> net/ipv6/udp.c:1001:DEFINE_PROTO_INUSE(udpv6)
> net/ipv6/udplite.c:43:DEFINE_PROTO_INUSE(udplitev6)
> net/ipv6/raw.c:1187:DEFINE_PROTO_INUSE(rawv6)
> net/ipv6/tcp_ipv6.c:2108:DEFINE_PROTO_INUSE(tcpv6)
> net/ipv4/udp.c:1477:DEFINE_PROTO_INUSE(udp)
> net/ipv4/udplite.c:47:DEFINE_PROTO_INUSE(udplite)
> net/ipv4/tcp_ipv4.c:2406:DEFINE_PROTO_INUSE(tcp)
> net/ipv4/raw.c:828:DEFINE_PROTO_INUSE(raw)
> net/sctp/socket.c:6461:DEFINE_PROTO_INUSE(sctp)
> net/sctp/socket.c:6495:DEFINE_PROTO_INUSE(sctpv6)
> include/net/sock.h:624:# define DEFINE_PROTO_INUSE(NAME) DEFINE_PCOUNTER(NAME)
> include/net/sock.h:644:# define DEFINE_PROTO_INUSE(NAME)
> 
> So pcounter_alloc()/pcounter_free() could just be deleted from pcounter.h

Well if you do that and if pcounters become helpers functions around
DEFINE_PERCPU memory then things are starting to get a bit more sensible. 
We can then remove the indirect function call.

Yes, a general wrapper around a DEFINE_PERCPU'ed counter could be a useful
thing to have.  It could use raw_smp_processor_id() and local_add().

Problem is the read side will still be far too inefficient for it to be
presentable as a general-purpose library - it really will need to do all
the batching which experience told us was needed in percpu_counters.

The read side should also present two interfaces: one which returns a +ve
or -ve value and one which returns a never-negative value.

> indirect functions calls are everywhere in kernel, network, fs, everywhere.

That doesn't make them fast.

> As soon we can put in 'struct pcounter' the address of a percpu variable, we 
> wont need anymore pointers to the pcounter_add()/getval() function.
> 
> mov  0(pcounter),%eax  # get the address of a percpuvar
> add  %edx,fs:%eax
> 
> This just need the percpu work done by SGI guys to be finished.

So...  I need to keep watching the tree to make sure that nobody actually
uses this code in ./lib/?

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Powered by blists - more mailing lists