From: Martin Schwidefsky <schwidefsky@de.ibm.com>
To: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>,
Hannes Frederic Sowa <hannes@stressinduktion.org>,
netdev@vger.kernel.org, dborkman@redhat.com,
darkjames-ws@darkjames.pl, Mircea Gherzan <mgherzan@gmail.com>,
Russell King <rmk+kernel@arm.linux.org.uk>,
Matt Evans <matt@ozlabs.org>
Subject: Re: [PATCH net] bpf: do not use reciprocal divide
Date: Wed, 15 Jan 2014 16:26:23 +0100 [thread overview]
Message-ID: <20140115162623.39472f8f@mschwide> (raw)
In-Reply-To: <1389795716.31367.333.camel@edumazet-glaptop2.roam.corp.google.com>
On Wed, 15 Jan 2014 06:21:56 -0800
Eric Dumazet <eric.dumazet@gmail.com> wrote:
> On Wed, 2014-01-15 at 11:51 +0100, Heiko Carstens wrote:
> > On Wed, Jan 15, 2014 at 09:13:22AM +0100, Martin Schwidefsky wrote:
> > > On Wed, 15 Jan 2014 09:00:07 +0100
> > > Heiko Carstens <heiko.carstens@de.ibm.com> wrote:
> > >
> > > > On Tue, Jan 14, 2014 at 11:02:41PM -0800, Eric Dumazet wrote:
> > > > > diff --git a/arch/s390/net/bpf_jit_comp.c b/arch/s390/net/bpf_jit_comp.c
> > > > > index 16871da37371..e349dc7d0992 100644
> > > > > --- a/arch/s390/net/bpf_jit_comp.c
> > > > > +++ b/arch/s390/net/bpf_jit_comp.c
> > > > > @@ -371,11 +371,11 @@ static int bpf_jit_insn(struct bpf_jit *jit, struct sock_filter *filter,
> > > > > /* dr %r4,%r12 */
> > > > > EMIT2(0x1d4c);
> > > > > break;
> > > > > - case BPF_S_ALU_DIV_K: /* A = reciprocal_divide(A, K) */
> > > > > - /* m %r4,<d(K)>(%r13) */
> > > > > - EMIT4_DISP(0x5c40d000, EMIT_CONST(K));
> > > > > - /* lr %r5,%r4 */
> > > > > - EMIT2(0x1854);
> > > > > + case BPF_S_ALU_DIV_K: /* A /= K */
> > > > > + /* lhi %r4,0 */
> > > > > + EMIT4(0xa7480000);
> > > > > + /* d %r4,<d(K)>(%r13) */
> > > > > + EMIT4_DISP(0x5d40d000, EMIT_CONST(K));
> > > > > break;
> > > >
> > > > The s390 part looks good.
> > >
> > > Does it? The divide instruction is signed, for the special
> > > case of K==1 this can now cause an exception if the quotient
> > > gets too large. We should add a check for K==1 and do nothing
> > > in this case. With a divisor of at least 2 the result will
> > > stay in the limit.
> >
> > Indeed. That's quite subtle.
>
> net/core/filter.c does :
>
> A /= K;
>
> Why is this working in generic code (if K == 1), not in s390 one ?
The C compiler naturally can to a u32/u32 division, it either uses
the "dlr" instruction which is unsigned, or uses a call to a function
to do the u32/u32 math. See the lovely code in arch/s390/lib/div64.c
for the kernel version of that code.
--
blue skies,
Martin.
"Reality continues to ruin my life." - Calvin.
next prev parent reply other threads:[~2014-01-15 15:26 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-01-13 21:42 [PATCH RFC] reciprocal_divide: correction/update of the algorithm Hannes Frederic Sowa
2014-01-14 18:02 ` Randy Dunlap
2014-01-15 15:02 ` Hannes Frederic Sowa
2014-01-14 18:07 ` Eric Dumazet
2014-01-14 19:22 ` Austin S Hemmelgarn
2014-01-14 19:50 ` Eric Dumazet
2014-01-14 20:10 ` Hannes Frederic Sowa
2014-01-14 20:53 ` Austin S Hemmelgarn
2014-01-14 22:45 ` Eric Dumazet
2014-01-14 23:25 ` Borislav Petkov
2014-01-15 2:51 ` Austin S. Hemmelgarn
2014-01-14 22:39 ` Hannes Frederic Sowa
2014-01-15 7:02 ` [PATCH net] bpf: do not use reciprocal divide Eric Dumazet
2014-01-15 7:28 ` David Miller
2014-01-15 7:39 ` Eric Dumazet
2014-01-15 8:00 ` Heiko Carstens
2014-01-15 8:13 ` Martin Schwidefsky
2014-01-15 10:51 ` Heiko Carstens
2014-01-15 14:21 ` Eric Dumazet
2014-01-15 14:25 ` Eric Dumazet
2014-01-15 14:50 ` [PATCH v2 " Eric Dumazet
2014-01-15 15:10 ` Matt Evans
2014-01-15 16:09 ` Eric Dumazet
2014-01-16 1:02 ` David Miller
2014-01-17 8:59 ` Heiko Carstens
2014-01-18 2:56 ` David Miller
2014-01-18 10:12 ` Heiko Carstens
2014-01-15 15:35 ` [PATCH " Martin Schwidefsky
2014-01-15 15:26 ` Martin Schwidefsky [this message]
2014-01-15 16:07 ` Eric Dumazet
2014-01-15 14:16 ` Eric Dumazet
2014-01-15 15:10 ` Heiko Carstens
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20140115162623.39472f8f@mschwide \
--to=schwidefsky@de.ibm.com \
--cc=darkjames-ws@darkjames.pl \
--cc=dborkman@redhat.com \
--cc=eric.dumazet@gmail.com \
--cc=hannes@stressinduktion.org \
--cc=heiko.carstens@de.ibm.com \
--cc=matt@ozlabs.org \
--cc=mgherzan@gmail.com \
--cc=netdev@vger.kernel.org \
--cc=rmk+kernel@arm.linux.org.uk \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).