From: Daniel Mack <daniel@zonque.org>
To: u-boot@lists.denx.de
Subject: [U-Boot] am335x: GPMC: reading speed with prefetch mode
Date: Thu, 19 Mar 2015 14:56:10 +0100 [thread overview]
Message-ID: <550AD57A.4070701@zonque.org> (raw)
In-Reply-To: <CAGm1_kvgLKxDt_Xy8tdVDfOk46sD0-ybycXZT3tUd-5-=mD8WA@mail.gmail.com>
Hi,
On 03/19/2015 02:41 PM, Yegor Yefremov wrote:
> I've got v2015.04-rc4 running on my custom am335x (600MHz) based
> board. My 8-bit NAND chip:
>
> [17.297793 0.004021] omap-gpmc 50000000.gpmc: GPMC revision 6.0
> [17.303850 0.006057] nand: device found, Manufacturer ID: 0x2c, Chip ID: 0xda
> [17.309706 0.005856] nand: Micron MT29F2G08ABAEAWP
> [17.312823 0.003117] nand: 256MiB, SLC, page size: 2048, OOB size: 64
> [17.317311 0.004488] nand: using OMAP_ECC_BCH8_CODE_HW ECC scheme
>
> I need to load about 17Mb FIT image from UBIFS partition. In Linux it
> takes about 7 seconds:
>
> # time cp /mnt/kernel-fit.itb /tmp/
> real 0m 7.12s
> user 0m 0.00s
> sys 0m 6.89s
>
> But U-Boot needs about twice the time:
On my boards, I didn't load the uImage through UBIFS but directly from
the raw mtdblock device. That might make a significant difference,
depending on how UBIFS is implemented in U-Boot. For performance tests,
I recommend you compare the numbers using 'dd' from the mtdblock device
under Linux, and 'nand read.i' from U-Boot.
Linux will, however, still be faster due to DMA, which is unsuable from
U-Boot due to the lack of interrupt handlers. But in my tests, enabling
the prefetch mode in U-Boot gave me a speed-up of roughly factor 2 IIRC.
Thanks,
Daniel
next prev parent reply other threads:[~2015-03-19 13:56 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-03-19 13:41 [U-Boot] am335x: GPMC: reading speed with prefetch mode Yegor Yefremov
2015-03-19 13:56 ` Daniel Mack [this message]
2015-03-19 15:13 ` Yegor Yefremov
2015-03-19 15:56 ` Daniel Mack
2015-03-20 11:24 ` Yegor Yefremov
2015-03-20 12:37 ` Roger Quadros
2015-03-20 13:37 ` Yegor Yefremov
2015-03-20 14:05 ` Roger Quadros
2015-03-20 15:09 ` Tony Lindgren
2015-03-20 14:54 ` Daniel Mack
2015-03-20 15:06 ` Yegor Yefremov
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=550AD57A.4070701@zonque.org \
--to=daniel@zonque.org \
--cc=u-boot@lists.denx.de \
/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