From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
To: Dan Williams <dan.j.williams@intel.com>
Cc: Ilya Yanok <yanok@emcraft.com>,
linux-raid@vger.kernel.org, linuxppc-dev@ozlabs.org, dzu@denx.de,
wd@denx.de
Subject: Re: [PATCH 01/11] async_tx: don't use src_list argument of async_xor() for dma addresses
Date: Sat, 15 Nov 2008 18:12:22 +1100 [thread overview]
Message-ID: <1226733142.7178.162.camel@pasglop> (raw)
In-Reply-To: <e9c3a7c20811141642q7608e642sd14b3021223a2d3c@mail.gmail.com>
On Fri, 2008-11-14 at 17:42 -0700, Dan Williams wrote:
> I don't like the stack space implications of this change. Especially
> for large arrays we will be carrying two 'src_cnt' size arrays on the
> stack, one from MD and one from async_tx. However, I think the
> current scheme of overwriting input parameters is pretty ugly.
Well, it's also broken :-) On a number of architectures, dma_addr_t can
be 64 bit while page * is 32 bit
> So, I
> want to benchmark the performance implications of adding a GFP_NOIO
> allocation here, with the idea being that if the allocation fails we
> can still fallback to the synchronous code path.
WARNING: multiple messages have this Message-ID (diff)
From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
To: Dan Williams <dan.j.williams@intel.com>
Cc: linux-raid@vger.kernel.org, linuxppc-dev@ozlabs.org, dzu@denx.de,
Ilya Yanok <yanok@emcraft.com>,
wd@denx.de
Subject: Re: [PATCH 01/11] async_tx: don't use src_list argument of async_xor() for dma addresses
Date: Sat, 15 Nov 2008 18:12:22 +1100 [thread overview]
Message-ID: <1226733142.7178.162.camel@pasglop> (raw)
In-Reply-To: <e9c3a7c20811141642q7608e642sd14b3021223a2d3c@mail.gmail.com>
On Fri, 2008-11-14 at 17:42 -0700, Dan Williams wrote:
> I don't like the stack space implications of this change. Especially
> for large arrays we will be carrying two 'src_cnt' size arrays on the
> stack, one from MD and one from async_tx. However, I think the
> current scheme of overwriting input parameters is pretty ugly.
Well, it's also broken :-) On a number of architectures, dma_addr_t can
be 64 bit while page * is 32 bit
> So, I
> want to benchmark the performance implications of adding a GFP_NOIO
> allocation here, with the idea being that if the allocation fails we
> can still fallback to the synchronous code path.
next prev parent reply other threads:[~2008-11-15 7:12 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-11-13 15:15 [RFC PATCH 00/11] md: support for asynchronous execution of RAID6 operations Ilya Yanok
2008-11-13 15:15 ` [PATCH 01/11] async_tx: don't use src_list argument of async_xor() for dma addresses Ilya Yanok
2008-11-15 0:42 ` Dan Williams
2008-11-15 0:42 ` Dan Williams
2008-11-15 7:12 ` Benjamin Herrenschmidt [this message]
2008-11-15 7:12 ` Benjamin Herrenschmidt
2008-11-13 15:15 ` [PATCH 02/11] async_tx: add support for asynchronous GF multiplication Ilya Yanok
2008-11-15 1:28 ` Dan Williams
2008-11-15 1:28 ` Dan Williams
2008-11-27 1:26 ` Re[2]: " Yuri Tikhonov
2008-11-27 1:26 ` Yuri Tikhonov
2008-11-28 21:18 ` Dan Williams
2008-11-28 21:18 ` Dan Williams
2008-11-13 15:15 ` [PATCH 03/11] async_tx: add support for asynchronous RAID6 recovery operations Ilya Yanok
2008-11-13 15:15 ` [PATCH 04/11] md: run stripe operations outside the lock Ilya Yanok
2008-11-13 15:15 ` [PATCH 05/11] md: common schedule_reconstruction for raid5/6 Ilya Yanok
2008-11-13 15:15 ` [PATCH 06/11] md: change handle_stripe_fill6 to work in asynchronous way Ilya Yanok
2008-11-13 15:16 ` [PATCH 07/11] md: rewrite handle_stripe_dirtying6 " Ilya Yanok
2008-11-13 15:16 ` [PATCH 08/11] md: asynchronous handle_parity_check6 Ilya Yanok
2008-11-13 15:16 ` [PATCH 09/11] md: change handle_stripe6 to work asynchronously Ilya Yanok
2008-11-13 15:16 ` [PATCH 10/11] md: remove unused functions Ilya Yanok
2008-11-13 15:16 ` [PATCH 11/11] ppc440spe-adma: ADMA driver for PPC440SP(e) systems Ilya Yanok
2008-11-13 16:03 ` Josh Boyer
2008-11-13 16:03 ` Josh Boyer
2008-11-13 17:50 ` Ilya Yanok
2008-11-13 17:50 ` Ilya Yanok
2008-11-13 17:54 ` Josh Boyer
2008-11-13 17:54 ` Josh Boyer
2008-12-09 1:08 ` Re[2]: " Yuri Tikhonov
2008-12-09 1:08 ` Yuri Tikhonov
2009-05-06 15:32 ` 440SPE ADMA driver Tirumala Reddy Marri
2009-05-06 15:32 ` Tirumala Reddy Marri
-- strict thread matches above, loose matches on Subject: below --
2008-12-08 21:55 [PATCH 01/11] async_tx: don't use src_list argument of async_xor() for dma addresses Yuri Tikhonov
2008-12-08 21:55 ` Yuri Tikhonov
2008-12-09 0:31 ` Dan Williams
2008-12-09 0:31 ` Dan Williams
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=1226733142.7178.162.camel@pasglop \
--to=benh@kernel.crashing.org \
--cc=dan.j.williams@intel.com \
--cc=dzu@denx.de \
--cc=linux-raid@vger.kernel.org \
--cc=linuxppc-dev@ozlabs.org \
--cc=wd@denx.de \
--cc=yanok@emcraft.com \
/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.