public inbox for u-boot@lists.denx.de
 help / color / mirror / Atom feed
From: Scott Wood <scottwood@freescale.com>
To: u-boot@lists.denx.de
Subject: [U-Boot] [PATCH] mmc:dcache: Cache line size aligned internal MMC buffers
Date: Mon, 29 Aug 2011 15:47:26 -0500	[thread overview]
Message-ID: <4E5BFADE.9000200@freescale.com> (raw)
In-Reply-To: <CAF6FioWhhqbVQWWPrusb69S2mpfBtoiazxc8x56bkogJmzXZ3g@mail.gmail.com>

On 08/29/2011 03:12 PM, Anton Staaf wrote:
> 1) Mikes's macro
> 
> #define DMA_ALIGN_SIZE(size) \
>        (((size) + CONFIG_SYS_CACHELINE_SIZE - 1)
> 
> #define DMA_DECLARE_BUFFER(type, name, size) \
>        void __##name[DMA_ALIGN_SIZE(size * sizeof(type))]; \
>        type * name = __##name & ~(CONFIG_SYS_CACHELINE_SIZE - 1));
> 
> DMA_DECLARE_BUFFER(int, buffer, 100);

This doesn't compile, and it tries to round the buffer down below its
starting point.

After fixing the more obvious issues, I get "error: initializer element
is not constant".

There might be no way to express and-by-constant as a relocation.

You could set the pointer at runtime, though, and remove some of the
macrification:

#define DMA_ALIGN_SIZE(size) \
	((size) + CONFIG_SYS_CACHELINE_SIZE - 1)
#define DMA_ALIGN_ADDR(addr) \
	(DMA_ALIGN_SIZE(addr) & (CONFIG_SYS_CACHELINE_SIZE - 1))

int buffer_unaligned[DMA_ALIGN_SIZE(100)];
int *buffer;

some_init_func()
{
	buffer = (int *)(DMA_ALIGN_ADDR((uintptr_t)buffer_unaligned));
}

> 3) Use GCC specific alignment attribute:
> 
> #define CACHLINE_ALIGNED __attribute__ ((aligned (CONFIG_SYS_CACHELINE_SIZE)))
> 
> int buffer[100] CACHELINE_ALIGNED;
> 
> Pros: The declaration of the buffer is even simpler and more obvious,
> no use of alloca at all.
> 
> Cons: This doesn't work in any version of GCC before October 2010.
> Meaning that it probably doesn't work in whatever compiler you're
> using.
> 
> 
> It's really too bad that this isn't a usable solution.  I suppose that
> we could switch to it at some point when we expect U-Boot to only be
> compiled by versions of GCC that support this.  By the way, the
> failure mode here is pretty bad.  If you compile the above code with
> an older GCC it will silently fail to align the variable.  :(

If the decision is made to depend on newer compilers, U-Boot could check
for it and #error out if the compiler is too old (possibly just in the
files that depend on this feature).

-Scott

  parent reply	other threads:[~2011-08-29 20:47 UTC|newest]

Thread overview: 55+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-08-19  9:25 [U-Boot] [PATCH] mmc:dcache: Cache line size aligned internal MMC buffers Lukasz Majewski
2011-08-19 13:57 ` Mike Frysinger
2011-08-19 15:28   ` Lukasz Majewski
2011-08-19 15:35     ` Mike Frysinger
2011-08-22  7:29       ` Lukasz Majewski
2011-08-22 16:08         ` Mike Frysinger
2011-08-22 16:42           ` Anton Staaf
2011-08-22 16:52             ` Marek Vasut
2011-08-22 17:17             ` Mike Frysinger
2011-08-22 18:15               ` Anton Staaf
2011-08-22 18:31                 ` Mike Frysinger
2011-08-22 18:57                   ` Anton Staaf
2011-08-23  9:19                     ` Lukasz Majewski
2011-08-23 17:00                       ` Anton Staaf
2011-08-23 17:30                       ` Mike Frysinger
2011-08-23 18:12                         ` Anton Staaf
2011-08-23 18:35                           ` Mike Frysinger
2011-08-23 18:36                           ` Mike Frysinger
2011-08-23 18:46                             ` Anton Staaf
2011-08-23 20:12                           ` Wolfgang Denk
2011-08-23 20:27                             ` Anton Staaf
2011-08-23 20:37                               ` Mike Frysinger
2011-08-23 21:06                                 ` Anton Staaf
2011-08-23 21:32                                   ` Mike Frysinger
2011-08-23 21:09                                 ` Wolfgang Denk
2011-08-23 21:32                                   ` Mike Frysinger
2011-08-23 21:48                                     ` Anton Staaf
2011-08-24 16:16                                       ` Mike Frysinger
2011-08-23 22:42                                     ` Wolfgang Denk
2011-08-24  3:00                                       ` Mike Frysinger
2011-08-24 10:07                                         ` Lukasz Majewski
2011-08-24 13:25                                           ` Wolfgang Denk
2011-08-24 14:31                                             ` Lukasz Majewski
2011-08-24 16:20                                             ` Mike Frysinger
2011-08-24 17:27                                             ` Anton Staaf
2011-08-24 18:06                                               ` Mike Frysinger
2011-08-24 18:12                                               ` Wolfgang Denk
2011-08-24 18:25                                                 ` Anton Staaf
2011-08-24 19:04                                                   ` Wolfgang Denk
2011-08-24 20:12                                                     ` Anton Staaf
2011-08-24 19:18                                                   ` Lukasz Majewski
2011-08-24 20:13                                                     ` Anton Staaf
2011-08-29 20:12                                                       ` Anton Staaf
2011-08-29 20:35                                                         ` Wolfgang Denk
2011-08-29 21:08                                                           ` Anton Staaf
2011-08-29 20:47                                                         ` Scott Wood [this message]
2011-08-29 20:58                                                           ` Anton Staaf
2011-08-29 21:23                                                             ` Scott Wood
2011-08-29 21:54                                                               ` Anton Staaf
2011-08-29 22:03                                                                 ` Scott Wood
2011-08-29 22:49                                                                   ` Anton Staaf
2011-08-29 23:01                                                                     ` Scott Wood
2011-08-29 23:05                                                                       ` Anton Staaf
2011-08-23 20:35                             ` Mike Frysinger
2011-08-23  8:42           ` Lukasz Majewski

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=4E5BFADE.9000200@freescale.com \
    --to=scottwood@freescale.com \
    --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