From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH] mm: do not print backtraces on GFP_ATOMIC failures Date: Tue, 21 Sep 2010 19:00:27 +0200 Message-ID: <1285088427.2617.723.camel@edumazet-laptop> References: <20100921121818.4745f038@annuminas.surriel.com> <20100921094638.9910add0.akpm@linux-foundation.org> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Cc: Rik van Riel , linux-mm@kvack.org, linux-kernel@vger.kernel.org, KOSAKI Motohiro , netdev@vger.kernel.org To: Andrew Morton Return-path: In-Reply-To: <20100921094638.9910add0.akpm@linux-foundation.org> Sender: owner-linux-mm@kvack.org List-Id: netdev.vger.kernel.org Le mardi 21 septembre 2010 =C3=A0 09:46 -0700, Andrew Morton a =C3=A9crit= : > On Tue, 21 Sep 2010 12:18:18 -0400 Rik van Riel wrote= : >=20 > > Atomic allocations cannot fall back to the page eviction code > > and are expected to fail. In fact, in some network intensive > > workloads, it is common to experience hundreds of GFP_ATOMIC > > allocation failures. > >=20 > > Printing out a backtrace for every one of those expected > > allocation failures accomplishes nothing good. At multi-gigabit > > network speeds with jumbo frames, a burst of allocation failure > > backtraces could even slow down the system. > >=20 > > We're better off not printing out backtraces on GFP_ATOMIC > > allocation failures. > >=20 > > Signed-off-by: Rik van Riel > >=20 > > diff --git a/include/linux/gfp.h b/include/linux/gfp.h > > index 975609c..5a0bddb 100644 > > --- a/include/linux/gfp.h > > +++ b/include/linux/gfp.h > > @@ -72,7 +72,7 @@ struct vm_area_struct; > > /* This equals 0, but use constants in case they ever change */ > > #define GFP_NOWAIT (GFP_ATOMIC & ~__GFP_HIGH) > > /* GFP_ATOMIC means both !wait (__GFP_WAIT not set) and use emergenc= y pool */ > > -#define GFP_ATOMIC (__GFP_HIGH) > > +#define GFP_ATOMIC (__GFP_HIGH | __GFP_NOWARN) > > #define GFP_NOIO (__GFP_WAIT) > > #define GFP_NOFS (__GFP_WAIT | __GFP_IO) > > #define GFP_KERNEL (__GFP_WAIT | __GFP_IO | __GFP_FS) >=20 > A much finer-tuned implementation would be to add __GFP_NOWARN just to > the networking call sites. I asked about this in June and it got > nixed: >=20 > http://www.spinics.net/lists/netdev/msg131965.html > -- Yes, I remember this particular report was useful to find and correct a bug. I dont know what to say. Being silent or verbose, it really depends on the context ? -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org