From: Matt Mackall <mpm@selenic.com>
To: David Miller <davem@davemloft.net>
Cc: lethal@linux-sh.org, penberg@cs.helsinki.fi,
herbert@gondor.apana.org.au, ken@codelabs.ch,
geert@linux-m68k.org, michael-dev@fami-braun.de,
linux-kernel@vger.kernel.org, linux-crypto@vger.kernel.org,
anemo@mba.ocn.ne.jp
Subject: Re: [BUG] SLOB breaks Crypto
Date: Tue, 18 May 2010 18:16:15 -0500 [thread overview]
Message-ID: <1274224575.11603.1248.camel@calx> (raw)
In-Reply-To: <20100518.154059.139137292.davem@davemloft.net>
On Tue, 2010-05-18 at 15:40 -0700, David Miller wrote:
> All of the CPP tests like the one used by linux/crypto.h are
> ludicrious. It should absolutely be not necessary for any code to
> duplicate this kind of calculation.
>
> Instead, this sequence should be in linux/slab.h, and be used
> universally by slab, slub, slob and anything that wants to know the
> allocators alignment guarentees.
Agreed. However, every arch should -also- set ARCH_KMALLOC_MINALIGN
appropriately so that we have documentation of the hardware requirements
on each platform.
> I don't even know of a 32-bit chip outside of x86 that doesn't
> potentially emit alignment requiring 64-bit memory operations for
> 64-bit objects. So what SLOB is doing with a different default is
> even more strange. And I bet you that even without the requirement,
> x86 runs faster with 64-bit alignment of 64-bit objects.
No doubt but SLOB explicitly trades faster for smaller.
--
Mathematics is the supreme nostalgia of our time.
WARNING: multiple messages have this Message-ID (diff)
From: Matt Mackall <mpm@selenic.com>
To: David Miller <davem@davemloft.net>
Cc: lethal@linux-sh.org, penberg@cs.helsinki.fi,
herbert@gondor.hengli.com.au, ken@codelabs.ch,
geert@linux-m68k.org, michael-dev@fami-braun.de,
linux-kernel@vger.kernel.org, linux-crypto@vger.kernel.org,
anemo@mba.ocn.ne.jp
Subject: Re: [BUG] SLOB breaks Crypto
Date: Tue, 18 May 2010 18:16:15 -0500 [thread overview]
Message-ID: <1274224575.11603.1248.camel@calx> (raw)
In-Reply-To: <20100518.154059.139137292.davem@davemloft.net>
On Tue, 2010-05-18 at 15:40 -0700, David Miller wrote:
> All of the CPP tests like the one used by linux/crypto.h are
> ludicrious. It should absolutely be not necessary for any code to
> duplicate this kind of calculation.
>
> Instead, this sequence should be in linux/slab.h, and be used
> universally by slab, slub, slob and anything that wants to know the
> allocators alignment guarentees.
Agreed. However, every arch should -also- set ARCH_KMALLOC_MINALIGN
appropriately so that we have documentation of the hardware requirements
on each platform.
> I don't even know of a 32-bit chip outside of x86 that doesn't
> potentially emit alignment requiring 64-bit memory operations for
> 64-bit objects. So what SLOB is doing with a different default is
> even more strange. And I bet you that even without the requirement,
> x86 runs faster with 64-bit alignment of 64-bit objects.
No doubt but SLOB explicitly trades faster for smaller.
--
Mathematics is the supreme nostalgia of our time.
next prev parent reply other threads:[~2010-05-18 23:16 UTC|newest]
Thread overview: 111+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-03-15 13:39 [BUG] SLOB breaks Crypto michael-dev
2010-03-18 16:30 ` Pekka Enberg
2010-03-18 21:24 ` michael-dev
2010-03-19 0:33 ` Herbert Xu
2010-05-14 14:50 ` Adrian-Ken Rueegsegger
2010-05-14 14:50 ` Adrian-Ken Rueegsegger
2010-05-17 16:17 ` Geert Uytterhoeven
2010-05-17 16:17 ` Geert Uytterhoeven
2010-05-17 21:50 ` Adrian-Ken Rueegsegger
2010-05-17 21:50 ` Adrian-Ken Rueegsegger
2010-05-17 22:37 ` Matt Mackall
2010-05-17 22:37 ` Matt Mackall
2010-05-18 8:17 ` Adrian-Ken Rueegsegger
2010-05-18 8:17 ` Adrian-Ken Rueegsegger
2010-05-18 10:27 ` Herbert Xu
2010-05-18 10:27 ` Herbert Xu
2010-05-18 14:02 ` Pekka Enberg
2010-05-18 14:02 ` Pekka Enberg
2010-05-18 19:06 ` Matt Mackall
2010-05-18 19:06 ` Matt Mackall
2010-05-18 19:25 ` David Miller
2010-05-18 19:25 ` David Miller
2010-05-18 19:33 ` Matt Mackall
2010-05-18 19:33 ` Matt Mackall
2010-05-18 20:59 ` David Miller
2010-05-18 20:59 ` David Miller
2010-05-18 21:15 ` Pekka Enberg
2010-05-18 21:15 ` Pekka Enberg
2010-05-18 21:20 ` David Miller
2010-05-18 21:20 ` David Miller
2010-05-18 22:35 ` Paul Mundt
2010-05-18 22:35 ` Paul Mundt
2010-05-18 22:37 ` Paul Mundt
2010-05-18 22:37 ` Paul Mundt
2010-05-19 15:19 ` Christoph Lameter
2010-05-19 19:56 ` David Miller
2010-05-18 22:40 ` David Miller
2010-05-18 22:40 ` David Miller
2010-05-18 23:10 ` Paul Mundt
2010-05-18 23:10 ` Paul Mundt
2010-05-18 23:16 ` Matt Mackall [this message]
2010-05-18 23:16 ` Matt Mackall
2010-05-19 5:53 ` Pekka Enberg
2010-05-19 5:53 ` Pekka Enberg
2010-05-18 23:20 ` David Woodhouse
2010-05-18 23:20 ` David Woodhouse
2010-05-19 1:05 ` Herbert Xu
2010-05-19 1:05 ` Herbert Xu
2010-05-19 7:14 ` David Woodhouse
2010-05-19 7:14 ` David Woodhouse
2010-05-19 7:45 ` Herbert Xu
2010-05-19 7:45 ` Herbert Xu
2010-05-19 11:32 ` Geert Uytterhoeven
2010-05-19 11:40 ` David Woodhouse
2010-05-19 11:50 ` Geert Uytterhoeven
2010-05-19 14:11 ` Matt Mackall
2010-05-19 19:33 ` David Miller
2010-05-20 3:38 ` FUJITA Tomonori
2010-05-20 3:46 ` David Miller
2010-05-19 12:02 ` FUJITA Tomonori
2010-05-19 12:19 ` David Woodhouse
2010-05-19 12:26 ` FUJITA Tomonori
2010-05-19 12:48 ` David Woodhouse
2010-05-19 10:58 ` David Woodhouse
2010-05-19 10:58 ` David Woodhouse
2010-05-19 11:01 ` [PATCH 1/4] mm: Move ARCH_SLAB_MINALIGN and ARCH_KMALLOC_MINALIGN to <linux/slab_def.h> David Woodhouse
2010-05-19 11:01 ` David Woodhouse
2010-05-19 13:30 ` Johannes Stezenbach
2010-05-19 13:30 ` Johannes Stezenbach
2010-05-19 18:03 ` Manfred Spraul
2010-05-19 18:10 ` David Woodhouse
2010-05-20 16:49 ` Manfred Spraul
2010-05-19 19:11 ` Pekka Enberg
2010-05-19 19:11 ` Pekka Enberg
2010-05-19 19:19 ` Christoph Lameter
2010-05-19 19:23 ` Pekka Enberg
2010-05-19 19:45 ` Christoph Lameter
2010-05-19 19:28 ` David Woodhouse
2010-05-19 11:01 ` [PATCH 2/4] mm: Move ARCH_SLAB_MINALIGN and ARCH_KMALLOC_MINALIGN to <linux/slob_def.h> David Woodhouse
2010-05-19 11:01 ` David Woodhouse
2010-05-19 11:02 ` [PATCH 3/4] mm: Move ARCH_SLAB_MINALIGN and ARCH_KMALLOC_MINALIGN to <linux/slub_def.h> David Woodhouse
2010-05-19 11:02 ` David Woodhouse
2010-05-19 11:02 ` [PATCH 4/4] crypto: Use ARCH_KMALLOC_MINALIGN for CRYPTO_MINALIGN now that it's exposed David Woodhouse
2010-05-19 11:02 ` David Woodhouse
2010-05-19 11:08 ` [BUG] SLOB breaks Crypto Pekka Enberg
2010-05-19 11:08 ` Pekka Enberg
2010-05-19 11:16 ` David Woodhouse
2010-05-19 11:16 ` David Woodhouse
2010-05-19 11:46 ` Herbert Xu
2010-05-19 11:46 ` Herbert Xu
2010-05-19 12:54 ` Pekka Enberg
2010-05-19 12:54 ` Pekka Enberg
2010-05-19 12:59 ` Herbert Xu
2010-05-19 12:59 ` Herbert Xu
2010-05-19 19:51 ` David Miller
2010-05-19 19:51 ` David Miller
2010-05-19 19:42 ` David Miller
2010-05-19 19:42 ` David Miller
2010-05-19 13:49 ` [PATCH 5/4] Provide __dma_aligned macro David Woodhouse
2010-05-19 13:49 ` David Woodhouse
2010-05-19 14:50 ` FUJITA Tomonori
2010-05-19 4:09 ` [BUG] SLOB breaks Crypto Herbert Xu
2010-05-19 4:09 ` Herbert Xu
2010-05-19 5:44 ` Pekka Enberg
2010-05-19 5:44 ` Pekka Enberg
2010-05-18 19:32 ` Adrian-Ken Rueegsegger
2010-05-18 19:32 ` Adrian-Ken Rueegsegger
2010-05-18 22:39 ` Herbert Xu
2010-05-18 22:39 ` Herbert Xu
2010-05-19 1:51 ` Herbert Xu
2010-05-19 1:51 ` Herbert Xu
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=1274224575.11603.1248.camel@calx \
--to=mpm@selenic.com \
--cc=anemo@mba.ocn.ne.jp \
--cc=davem@davemloft.net \
--cc=geert@linux-m68k.org \
--cc=herbert@gondor.apana.org.au \
--cc=ken@codelabs.ch \
--cc=lethal@linux-sh.org \
--cc=linux-crypto@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=michael-dev@fami-braun.de \
--cc=penberg@cs.helsinki.fi \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.